Last active
July 2, 2021 13:44
-
-
Save Lord-V15/bca7dc3dbe11cada97f255c6538de897 to your computer and use it in GitHub Desktop.
NLP Engineer - VIBHANSH
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| { | |
| "nbformat": 4, | |
| "nbformat_minor": 0, | |
| "metadata": { | |
| "colab": { | |
| "name": "NLP Engineer - VIBHANSH", | |
| "provenance": [], | |
| "collapsed_sections": [], | |
| "include_colab_link": true | |
| }, | |
| "kernelspec": { | |
| "name": "python3", | |
| "display_name": "Python 3" | |
| }, | |
| "language_info": { | |
| "name": "python" | |
| }, | |
| "accelerator": "GPU" | |
| }, | |
| "cells": [ | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "view-in-github", | |
| "colab_type": "text" | |
| }, | |
| "source": [ | |
| "<a href=\"https://colab.research.google.com/gist/Lord-V15/bca7dc3dbe11cada97f255c6538de897/nlp-engineer-_-technical-selection-round.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "1j1iQo5365SN" | |
| }, | |
| "source": [ | |
| "" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "e0OYDAB7XTCW" | |
| }, | |
| "source": [ | |
| "#TEXTIFY AI NLP ENGINEER_TECHNICAL SELECTION ROUND\n", | |
| "\n", | |
| "#HELLO EVERYONE\n", | |
| "\n", | |
| "#Thank you for taking your time to work on these problem sets.\n", | |
| "\n", | |
| "#The aim of this problem set is to evaluate your skills in the following domains:\n", | |
| "# -Python programming basics (list comprehension, iterations, looping, if-else statements, etc )\n", | |
| "# -Web scrapping libraries (BeautifulSoup, requests)\n", | |
| "# -Data structures library (Pandas)\n", | |
| "# -NLP libraries (SpaCy, NLTK)\n", | |
| "# -ML/DL library (scikit-learn or TensorFlow)\n", | |
| "# -Word embedding models (word2vec or GloVe)\n", | |
| "\n", | |
| "#Please follow step-by-step process as mentioned below, and submit your solutions before 20th July.\n", | |
| "\n", | |
| "#Problem set difficulty -> Intermediate ; Time Requirement-> 4 to 6 hours (experienced developer), 2 to 3 days (Learners)\n", | |
| "\n", | |
| "#In case of any difficulty in accessing the required files, please contact the team at textifyai@gmail.com\n", | |
| "\n", | |
| "#Applicant's name -> VIBHANSH GUPTA" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "eqKy9jp1BmVv" | |
| }, | |
| "source": [ | |
| "**STEP 1**- Getting Started" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "57r38-36Bf-P", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 204 | |
| }, | |
| "outputId": "6e6b4c11-17b6-47c3-f638-fbe69ecc2702" | |
| }, | |
| "source": [ | |
| "#-Download the dataset (Essay_samples.csv) from your email.\n", | |
| "#-Copy this notebook into your own google Colab account.\n", | |
| "#-Upload the dataset on the notebook as below\n", | |
| "\n", | |
| "# from google.colab import files\n", | |
| "# uploaded = files.upload()\n", | |
| "\n", | |
| "#SKIPPING THE ABOVE CODE BECAUSE I ALREADY HAVE THE FILE UPLOADED DIRECTLY !\n", | |
| "\n", | |
| "import pandas as pd\n", | |
| "df_essay = pd.read_csv('Essay_samples - Sheet1.csv')\n", | |
| "\n", | |
| "df_essay.head()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/html": [ | |
| "<div>\n", | |
| "<style scoped>\n", | |
| " .dataframe tbody tr th:only-of-type {\n", | |
| " vertical-align: middle;\n", | |
| " }\n", | |
| "\n", | |
| " .dataframe tbody tr th {\n", | |
| " vertical-align: top;\n", | |
| " }\n", | |
| "\n", | |
| " .dataframe thead th {\n", | |
| " text-align: right;\n", | |
| " }\n", | |
| "</style>\n", | |
| "<table border=\"1\" class=\"dataframe\">\n", | |
| " <thead>\n", | |
| " <tr style=\"text-align: right;\">\n", | |
| " <th></th>\n", | |
| " <th>Essay Text</th>\n", | |
| " </tr>\n", | |
| " </thead>\n", | |
| " <tbody>\n", | |
| " <tr>\n", | |
| " <th>0</th>\n", | |
| " <td>THE ALARM CLOCK IS, TO MANY high school studen...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>1</th>\n", | |
| " <td>I HAVE ALWAYS BEEN A MATH-SCIENCE girl. I sigh...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>2</th>\n", | |
| " <td>WHEN I WAS FOUR YEARS OLD, I fell in love. It ...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>3</th>\n", | |
| " <td>THIS SUMMER, I WENT TO THE governor’s Honors P...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>4</th>\n", | |
| " <td>THIS PAST SUMMER I HAD THE opportunity to part...</td>\n", | |
| " </tr>\n", | |
| " </tbody>\n", | |
| "</table>\n", | |
| "</div>" | |
| ], | |
| "text/plain": [ | |
| " Essay Text\n", | |
| "0 THE ALARM CLOCK IS, TO MANY high school studen...\n", | |
| "1 I HAVE ALWAYS BEEN A MATH-SCIENCE girl. I sigh...\n", | |
| "2 WHEN I WAS FOUR YEARS OLD, I fell in love. It ...\n", | |
| "3 THIS SUMMER, I WENT TO THE governor’s Honors P...\n", | |
| "4 THIS PAST SUMMER I HAD THE opportunity to part..." | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 4 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "XJaD7z-JYve8" | |
| }, | |
| "source": [ | |
| "**STEP 2** - Web Scrapping" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "v5U1KArkVPev" | |
| }, | |
| "source": [ | |
| "## Extracting News from BBC requires the following :\n", | |
| "\n", | |
| "\n", | |
| "* URLs of Articles\n", | |
| "* Source Element of Title\n", | |
| "* Source Element of Content\n", | |
| "\n", | |
| "*Once done, data will be saved using Pandas*" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "TqgEXWRxYzfC", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "outputId": "c83cf708-b3e8-4dbd-fff1-3e80971a7c82" | |
| }, | |
| "source": [ | |
| "#-Use BeautifulSoup library to scrape 50 articles from (www.bbc.com).\n", | |
| "#-We will then use the scrapped articles to create a hybrid database.\n", | |
| "\n", | |
| "! pip install beautifulsoup4\n", | |
| "! pip install requests\n", | |
| "\n", | |
| "import requests\n", | |
| "from bs4 import BeautifulSoup\n", | |
| "\n", | |
| "url = 'https://www.bbc.com/sport/football' # I had to change the URL because I'm a massive Football fan :)\n", | |
| "response = requests.get(url)\n", | |
| "data = response.text\n", | |
| "soup = BeautifulSoup(data, 'html.parser')\n", | |
| "\n", | |
| "data = response.content # Content of response\n", | |
| "soup = BeautifulSoup(data, \"html.parser\")\n", | |
| "links = []\n", | |
| "for i in soup.find_all('div',{'class':'gel-layout__item gel-1/3@m gel-1/4@xxl sp-o-keyline sp-o-no-keyline@m'}): # Searching for news feed div class\n", | |
| " link = i.find('a',href=True) # Extracting attached link\n", | |
| " if str(link['href'][:6]) == '/sport' : # Making sure no ad links are extracted\n", | |
| " links.append( 'https://www.bbc.com'+str(link['href']) )\n", | |
| "print(links)\n", | |
| "\n", | |
| "texts= []\n", | |
| "titles= []\n", | |
| "for url in links :\n", | |
| " article = requests.get(url)\n", | |
| " soup = BeautifulSoup(article.content, \"html.parser\")\n", | |
| " title = soup.find(class_=['gel-trafalgar-bold qa-story-headline gs-u-mv+','gs-u-mb']).text # Extracting article heading\n", | |
| " print(f\"\\n TITLE : {title}\")\n", | |
| " titles.append(title)\n", | |
| "\n", | |
| "# There are 2 different type of content div classes so we will use them in conjunction below\n", | |
| " div = soup.find_all(\"div\", class_=[\"qa-story-body story-body gel-pica gel-10/12@m gel-7/8@l gs-u-ml0@l\", \"qa-story-body story-body gel-pica gel-10/12@m gel-7/8@l gs-u-ml0@l gs-u-pb++\"])\n", | |
| " for i in div:\n", | |
| " body = i.find_all(\"p\")\n", | |
| " text = ''\n", | |
| " for para in body :\n", | |
| " text+=f\" {para.text}\"\n", | |
| " texts.append(text)\n" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Requirement already satisfied: beautifulsoup4 in /usr/local/lib/python3.7/dist-packages (4.6.3)\n", | |
| "Requirement already satisfied: requests in /usr/local/lib/python3.7/dist-packages (2.23.0)\n", | |
| "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests) (3.0.4)\n", | |
| "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests) (1.24.3)\n", | |
| "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests) (2.10)\n", | |
| "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests) (2021.5.30)\n", | |
| "['https://www.bbc.com/sport/football/51198657', 'https://www.bbc.com/sport/football/51198650', 'https://www.bbc.com/sport/football/57680444', 'https://www.bbc.com/sport/football/57679283', 'https://www.bbc.com/sport/football/57692813', 'https://www.bbc.com/sport/football/57674033', 'https://www.bbc.com/sport/football/57673841', 'https://www.bbc.com/sport/football/57676732']\n", | |
| "\n", | |
| " TITLE : FRI 02 Jul 2021European Championship - Quarter-final\n", | |
| "\n", | |
| " TITLE : FRI 02 Jul 2021European Championship - Quarter-final\n", | |
| "\n", | |
| " TITLE : Switzerland v Spain: Swiss striker Haris Seferovic writing new chapter after difficult relationship with Spain\n", | |
| "\n", | |
| " TITLE : Euro 2020: Lawro takes on BBC Radio 5 Live pundits to predict quarter-final matches\n", | |
| "\n", | |
| " TITLE : Billy Gilmour: Norwich City sign Chelsea and Scotland midfielder on loan for season\n", | |
| "\n", | |
| " TITLE : Jadon Sancho: 'A talented trailblazer and hybrid of the modern age'\n", | |
| "\n", | |
| " TITLE : Rafael Benitez's love of Liverpool is not defined by a tribal club divide - Guillem Balague\n", | |
| "\n", | |
| " TITLE : Nuno Espirito Santo: Can Tottenham manager give the club a new way forward?\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 204 | |
| }, | |
| "id": "fiyfd3W3Gdfn", | |
| "outputId": "795a356f-6c8c-411f-babe-880f864b3a42" | |
| }, | |
| "source": [ | |
| "#-Create a new pandas table and add the essay database and the scrapped articles.\n", | |
| "dict_data = {'headline':titles, 'url':links, 'content':texts} \n", | |
| "df_scraped = pd.DataFrame(dict_data) #converting dict to pandas dataframe\n", | |
| "df_scraped.head()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/html": [ | |
| "<div>\n", | |
| "<style scoped>\n", | |
| " .dataframe tbody tr th:only-of-type {\n", | |
| " vertical-align: middle;\n", | |
| " }\n", | |
| "\n", | |
| " .dataframe tbody tr th {\n", | |
| " vertical-align: top;\n", | |
| " }\n", | |
| "\n", | |
| " .dataframe thead th {\n", | |
| " text-align: right;\n", | |
| " }\n", | |
| "</style>\n", | |
| "<table border=\"1\" class=\"dataframe\">\n", | |
| " <thead>\n", | |
| " <tr style=\"text-align: right;\">\n", | |
| " <th></th>\n", | |
| " <th>headline</th>\n", | |
| " <th>url</th>\n", | |
| " <th>content</th>\n", | |
| " </tr>\n", | |
| " </thead>\n", | |
| " <tbody>\n", | |
| " <tr>\n", | |
| " <th>0</th>\n", | |
| " <td>FRI 02 Jul 2021European Championship - Quarter...</td>\n", | |
| " <td>https://www.bbc.com/sport/football/51198657</td>\n", | |
| " <td>Belgium have fitness doubts over key duo Kevi...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>1</th>\n", | |
| " <td>FRI 02 Jul 2021European Championship - Quarter...</td>\n", | |
| " <td>https://www.bbc.com/sport/football/51198650</td>\n", | |
| " <td>Switzerland are \"hungry for more\" after shock...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>2</th>\n", | |
| " <td>Switzerland v Spain: Swiss striker Haris Sefer...</td>\n", | |
| " <td>https://www.bbc.com/sport/football/57680444</td>\n", | |
| " <td>It was supposed to be a night of fun for Hari...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>3</th>\n", | |
| " <td>Euro 2020: Lawro takes on BBC Radio 5 Live pun...</td>\n", | |
| " <td>https://www.bbc.com/sport/football/57679283</td>\n", | |
| " <td>BBC Sport football expert Mark Lawrenson is t...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>4</th>\n", | |
| " <td>Billy Gilmour: Norwich City sign Chelsea and S...</td>\n", | |
| " <td>https://www.bbc.com/sport/football/57692813</td>\n", | |
| " <td>Norwich City have signed Scotland midfielder ...</td>\n", | |
| " </tr>\n", | |
| " </tbody>\n", | |
| "</table>\n", | |
| "</div>" | |
| ], | |
| "text/plain": [ | |
| " headline ... content\n", | |
| "0 FRI 02 Jul 2021European Championship - Quarter... ... Belgium have fitness doubts over key duo Kevi...\n", | |
| "1 FRI 02 Jul 2021European Championship - Quarter... ... Switzerland are \"hungry for more\" after shock...\n", | |
| "2 Switzerland v Spain: Swiss striker Haris Sefer... ... It was supposed to be a night of fun for Hari...\n", | |
| "3 Euro 2020: Lawro takes on BBC Radio 5 Live pun... ... BBC Sport football expert Mark Lawrenson is t...\n", | |
| "4 Billy Gilmour: Norwich City sign Chelsea and S... ... Norwich City have signed Scotland midfielder ...\n", | |
| "\n", | |
| "[5 rows x 3 columns]" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 6 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "WYCYGeJzPY4Z", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 419 | |
| }, | |
| "outputId": "69a290ba-384a-4ea6-94c7-0b08075c5cf7" | |
| }, | |
| "source": [ | |
| "#df_new = df_essay + df_scraped_articles\n", | |
| "df_new = pd.concat([df_essay, pd.DataFrame({'Essay Text':texts})],axis=0,ignore_index=True) #axis=0 to concat rows and index is also reset\n", | |
| "df_new" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/html": [ | |
| "<div>\n", | |
| "<style scoped>\n", | |
| " .dataframe tbody tr th:only-of-type {\n", | |
| " vertical-align: middle;\n", | |
| " }\n", | |
| "\n", | |
| " .dataframe tbody tr th {\n", | |
| " vertical-align: top;\n", | |
| " }\n", | |
| "\n", | |
| " .dataframe thead th {\n", | |
| " text-align: right;\n", | |
| " }\n", | |
| "</style>\n", | |
| "<table border=\"1\" class=\"dataframe\">\n", | |
| " <thead>\n", | |
| " <tr style=\"text-align: right;\">\n", | |
| " <th></th>\n", | |
| " <th>Essay Text</th>\n", | |
| " </tr>\n", | |
| " </thead>\n", | |
| " <tbody>\n", | |
| " <tr>\n", | |
| " <th>0</th>\n", | |
| " <td>THE ALARM CLOCK IS, TO MANY high school studen...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>1</th>\n", | |
| " <td>I HAVE ALWAYS BEEN A MATH-SCIENCE girl. I sigh...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>2</th>\n", | |
| " <td>WHEN I WAS FOUR YEARS OLD, I fell in love. It ...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>3</th>\n", | |
| " <td>THIS SUMMER, I WENT TO THE governor’s Honors P...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>4</th>\n", | |
| " <td>THIS PAST SUMMER I HAD THE opportunity to part...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>...</th>\n", | |
| " <td>...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>102</th>\n", | |
| " <td>BBC Sport football expert Mark Lawrenson is t...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>103</th>\n", | |
| " <td>Norwich City have signed Scotland midfielder ...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>104</th>\n", | |
| " <td>The 'cultural reboot' Manchester United let i...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>105</th>\n", | |
| " <td>Apart from being Spanish and involved in foot...</td>\n", | |
| " </tr>\n", | |
| " <tr>\n", | |
| " <th>106</th>\n", | |
| " <td>Tottenham fans could be forgiven for feeling ...</td>\n", | |
| " </tr>\n", | |
| " </tbody>\n", | |
| "</table>\n", | |
| "<p>107 rows × 1 columns</p>\n", | |
| "</div>" | |
| ], | |
| "text/plain": [ | |
| " Essay Text\n", | |
| "0 THE ALARM CLOCK IS, TO MANY high school studen...\n", | |
| "1 I HAVE ALWAYS BEEN A MATH-SCIENCE girl. I sigh...\n", | |
| "2 WHEN I WAS FOUR YEARS OLD, I fell in love. It ...\n", | |
| "3 THIS SUMMER, I WENT TO THE governor’s Honors P...\n", | |
| "4 THIS PAST SUMMER I HAD THE opportunity to part...\n", | |
| ".. ...\n", | |
| "102 BBC Sport football expert Mark Lawrenson is t...\n", | |
| "103 Norwich City have signed Scotland midfielder ...\n", | |
| "104 The 'cultural reboot' Manchester United let i...\n", | |
| "105 Apart from being Spanish and involved in foot...\n", | |
| "106 Tottenham fans could be forgiven for feeling ...\n", | |
| "\n", | |
| "[107 rows x 1 columns]" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 7 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "aJ4Si0r7eRzn" | |
| }, | |
| "source": [ | |
| "**STEP 3** - Spacy dependency parsing" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "OKtgVowUQGja", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "outputId": "a0fc4829-0fff-4ce5-9634-ab82e499f28b" | |
| }, | |
| "source": [ | |
| "!spacy download en_core_web_sm" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "2021-07-02 10:34:32.474827: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcudart.so.11.0\n", | |
| "Collecting en-core-web-sm==3.0.0\n", | |
| "\u001b[?25l Downloading https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.0.0/en_core_web_sm-3.0.0-py3-none-any.whl (13.7MB)\n", | |
| "\u001b[K |████████████████████████████████| 13.7MB 225kB/s \n", | |
| "\u001b[?25hRequirement already satisfied: spacy<3.1.0,>=3.0.0 in /usr/local/lib/python3.7/dist-packages (from en-core-web-sm==3.0.0) (3.0.6)\n", | |
| "Requirement already satisfied: jinja2 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2.11.3)\n", | |
| "Requirement already satisfied: catalogue<2.1.0,>=2.0.3 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2.0.4)\n", | |
| "Requirement already satisfied: blis<0.8.0,>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (0.4.1)\n", | |
| "Requirement already satisfied: spacy-legacy<3.1.0,>=3.0.4 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (3.0.6)\n", | |
| "Requirement already satisfied: wasabi<1.1.0,>=0.8.1 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (0.8.2)\n", | |
| "Requirement already satisfied: preshed<3.1.0,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (3.0.5)\n", | |
| "Requirement already satisfied: murmurhash<1.1.0,>=0.28.0 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (1.0.5)\n", | |
| "Requirement already satisfied: thinc<8.1.0,>=8.0.3 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (8.0.7)\n", | |
| "Requirement already satisfied: srsly<3.0.0,>=2.4.1 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2.4.1)\n", | |
| "Requirement already satisfied: typing-extensions<4.0.0.0,>=3.7.4; python_version < \"3.8\" in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (3.7.4.3)\n", | |
| "Requirement already satisfied: cymem<2.1.0,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2.0.5)\n", | |
| "Requirement already satisfied: typer<0.4.0,>=0.3.0 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (0.3.2)\n", | |
| "Requirement already satisfied: requests<3.0.0,>=2.13.0 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2.23.0)\n", | |
| "Requirement already satisfied: numpy>=1.15.0 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (1.19.5)\n", | |
| "Requirement already satisfied: pathy>=0.3.5 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (0.6.0)\n", | |
| "Requirement already satisfied: pydantic<1.8.0,>=1.7.1 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (1.7.4)\n", | |
| "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (20.9)\n", | |
| "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (57.0.0)\n", | |
| "Requirement already satisfied: tqdm<5.0.0,>=4.38.0 in /usr/local/lib/python3.7/dist-packages (from spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (4.41.1)\n", | |
| "Requirement already satisfied: MarkupSafe>=0.23 in /usr/local/lib/python3.7/dist-packages (from jinja2->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2.0.1)\n", | |
| "Requirement already satisfied: zipp>=0.5; python_version < \"3.8\" in /usr/local/lib/python3.7/dist-packages (from catalogue<2.1.0,>=2.0.3->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (3.4.1)\n", | |
| "Requirement already satisfied: click<7.2.0,>=7.1.1 in /usr/local/lib/python3.7/dist-packages (from typer<0.4.0,>=0.3.0->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (7.1.2)\n", | |
| "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2021.5.30)\n", | |
| "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2.10)\n", | |
| "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (3.0.4)\n", | |
| "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (1.24.3)\n", | |
| "Requirement already satisfied: smart-open<6.0.0,>=5.0.0 in /usr/local/lib/python3.7/dist-packages (from pathy>=0.3.5->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (5.1.0)\n", | |
| "Requirement already satisfied: pyparsing>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging>=20.0->spacy<3.1.0,>=3.0.0->en-core-web-sm==3.0.0) (2.4.7)\n", | |
| "Installing collected packages: en-core-web-sm\n", | |
| " Found existing installation: en-core-web-sm 2.2.5\n", | |
| " Uninstalling en-core-web-sm-2.2.5:\n", | |
| " Successfully uninstalled en-core-web-sm-2.2.5\n", | |
| "Successfully installed en-core-web-sm-3.0.0\n", | |
| "\u001b[38;5;2m✔ Download and installation successful\u001b[0m\n", | |
| "You can now load the package via spacy.load('en_core_web_sm')\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "lGULbqAjmKrm", | |
| "outputId": "7fdcd926-4eb8-4619-d15f-309fcd1bc366" | |
| }, | |
| "source": [ | |
| "!pip install pytextrank" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Requirement already satisfied: pytextrank in /usr/local/lib/python3.7/dist-packages (3.1.1)\n", | |
| "Requirement already satisfied: graphviz>=0.13 in /usr/local/lib/python3.7/dist-packages (from pytextrank) (0.16)\n", | |
| "Requirement already satisfied: spacy>=3.0 in /usr/local/lib/python3.7/dist-packages (from pytextrank) (3.0.6)\n", | |
| "Requirement already satisfied: icecream>=2.1 in /usr/local/lib/python3.7/dist-packages (from pytextrank) (2.1.1)\n", | |
| "Requirement already satisfied: networkx>=2.0 in /usr/local/lib/python3.7/dist-packages (from pytextrank) (2.5.1)\n", | |
| "Requirement already satisfied: blis<0.8.0,>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (0.4.1)\n", | |
| "Requirement already satisfied: preshed<3.1.0,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (3.0.5)\n", | |
| "Requirement already satisfied: requests<3.0.0,>=2.13.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (2.23.0)\n", | |
| "Requirement already satisfied: thinc<8.1.0,>=8.0.3 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (8.0.7)\n", | |
| "Requirement already satisfied: pathy>=0.3.5 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (0.6.0)\n", | |
| "Requirement already satisfied: murmurhash<1.1.0,>=0.28.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (1.0.5)\n", | |
| "Requirement already satisfied: typing-extensions<4.0.0.0,>=3.7.4; python_version < \"3.8\" in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (3.7.4.3)\n", | |
| "Requirement already satisfied: typer<0.4.0,>=0.3.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (0.3.2)\n", | |
| "Requirement already satisfied: catalogue<2.1.0,>=2.0.3 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (2.0.4)\n", | |
| "Requirement already satisfied: wasabi<1.1.0,>=0.8.1 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (0.8.2)\n", | |
| "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (57.0.0)\n", | |
| "Requirement already satisfied: tqdm<5.0.0,>=4.38.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (4.41.1)\n", | |
| "Requirement already satisfied: spacy-legacy<3.1.0,>=3.0.4 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (3.0.6)\n", | |
| "Requirement already satisfied: srsly<3.0.0,>=2.4.1 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (2.4.1)\n", | |
| "Requirement already satisfied: cymem<2.1.0,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (2.0.5)\n", | |
| "Requirement already satisfied: pydantic<1.8.0,>=1.7.1 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (1.7.4)\n", | |
| "Requirement already satisfied: numpy>=1.15.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (1.19.5)\n", | |
| "Requirement already satisfied: jinja2 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (2.11.3)\n", | |
| "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.7/dist-packages (from spacy>=3.0->pytextrank) (20.9)\n", | |
| "Requirement already satisfied: colorama>=0.3.9 in /usr/local/lib/python3.7/dist-packages (from icecream>=2.1->pytextrank) (0.4.4)\n", | |
| "Requirement already satisfied: asttokens>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from icecream>=2.1->pytextrank) (2.0.5)\n", | |
| "Requirement already satisfied: executing>=0.3.1 in /usr/local/lib/python3.7/dist-packages (from icecream>=2.1->pytextrank) (0.6.0)\n", | |
| "Requirement already satisfied: pygments>=2.2.0 in /usr/local/lib/python3.7/dist-packages (from icecream>=2.1->pytextrank) (2.6.1)\n", | |
| "Requirement already satisfied: decorator<5,>=4.3 in /usr/local/lib/python3.7/dist-packages (from networkx>=2.0->pytextrank) (4.4.2)\n", | |
| "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy>=3.0->pytextrank) (2021.5.30)\n", | |
| "Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy>=3.0->pytextrank) (1.24.3)\n", | |
| "Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy>=3.0->pytextrank) (3.0.4)\n", | |
| "Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3.0.0,>=2.13.0->spacy>=3.0->pytextrank) (2.10)\n", | |
| "Requirement already satisfied: smart-open<6.0.0,>=5.0.0 in /usr/local/lib/python3.7/dist-packages (from pathy>=0.3.5->spacy>=3.0->pytextrank) (5.1.0)\n", | |
| "Requirement already satisfied: click<7.2.0,>=7.1.1 in /usr/local/lib/python3.7/dist-packages (from typer<0.4.0,>=0.3.0->spacy>=3.0->pytextrank) (7.1.2)\n", | |
| "Requirement already satisfied: zipp>=0.5; python_version < \"3.8\" in /usr/local/lib/python3.7/dist-packages (from catalogue<2.1.0,>=2.0.3->spacy>=3.0->pytextrank) (3.4.1)\n", | |
| "Requirement already satisfied: MarkupSafe>=0.23 in /usr/local/lib/python3.7/dist-packages (from jinja2->spacy>=3.0->pytextrank) (2.0.1)\n", | |
| "Requirement already satisfied: pyparsing>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging>=20.0->spacy>=3.0->pytextrank) (2.4.7)\n", | |
| "Requirement already satisfied: six in /usr/local/lib/python3.7/dist-packages (from asttokens>=2.0.1->icecream>=2.1->pytextrank) (1.15.0)\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "NK5KwYTEeWMB" | |
| }, | |
| "source": [ | |
| "#import Spacy library\n", | |
| "\n", | |
| "import spacy\n", | |
| "\n", | |
| "nlp = spacy.load('en_core_web_sm')\n", | |
| "\n", | |
| "#Use .split(\"\\n\") to split the essays into paragraphs.\n", | |
| "paragraphs =[ i.split(\"\\n\") for i in df_new['Essay Text'] ]\n" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 105 | |
| }, | |
| "id": "R-T1U4v6qJFY", | |
| "outputId": "b1428636-9240-4928-fec2-5b19d6bce696" | |
| }, | |
| "source": [ | |
| "paragraphs[0][1] #example result of splitting" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "application/vnd.google.colaboratory.intrinsic+json": { | |
| "type": "string" | |
| }, | |
| "text/plain": [ | |
| "'It isn’t because the only thing they desire is to sleep a few extra hours, as many would presume. no, these kids are groggy and irritable because they are waking up to what they think will be another horribly boring day of school. If one of these foolish Sallys or Joes were, say, sleeping comfortably on a Saturday morning, I could certainly see something different happening. A beautiful breakfast of tantalizing vittles—eggs, hash browns, and the like—would be ready and waiting for them on their kitchen tables. But the scrumptious delight to outshine them all would be a slab of bacon, piled proudly for the taking. It would be that wafting, wondrous bacon smell that would draw dear, sweet Sally abruptly from her slumber—long before an alarm clock has the chance to pierce the air. '" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 36 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "S5GiNgKbWoAb" | |
| }, | |
| "source": [ | |
| "## Pre-Processing\n", | |
| "Text needs to be cleaned for usage in the upcoming steps. The function is created here and called at multiple locations." | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "D0KyOt682lNq" | |
| }, | |
| "source": [ | |
| "import nltk\n", | |
| "import re\n", | |
| "from nltk.corpus import stopwords\n", | |
| "from nltk.stem import SnowballStemmer\n", | |
| "\n", | |
| "nltk.download('stopwords')\n", | |
| "\n", | |
| "stop_words = stopwords.words(\"english\") # List of words that don't contribute to the semantics\n", | |
| "stemmer = SnowballStemmer(\"english\") # Stemming to cut down the ends of words\n", | |
| "\n", | |
| "TEXT_CLEANING_RE = \"@\\S+|https?:\\S+|http?:\\S|[^A-Za-z0-9]+\" # Regular Expression for removing any links or illegible words\n", | |
| "\n", | |
| "def preprocess(text, stem=False):\n", | |
| " \n", | |
| " text = re.sub(TEXT_CLEANING_RE, ' ', str(text).lower()).strip() #Lower and clean and strip words\n", | |
| " tokens = []\n", | |
| " for token in text.split():\n", | |
| " if token not in stop_words: # If text is in stopword list, then ignore that text\n", | |
| " if stem:\n", | |
| " tokens.append(stemmer.stem(token)) # Stemming actually led to bad results so I set it to False by default\n", | |
| " else:\n", | |
| " tokens.append(token)\n", | |
| " return \" \".join(tokens) # Join text to recreate the sentences \n", | |
| "\n", | |
| "sentences_list = []\n", | |
| "for para in paragraphs :\n", | |
| " for sentences in para :\n", | |
| " temp = preprocess(sentences) # Adding all sentences of the corpus in one list\n", | |
| " sentences_list.append(temp)" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "S4yj5Ah0W9Ym" | |
| }, | |
| "source": [ | |
| "## **Reason for using PyTextRank**\n", | |
| "I am a contributor for this algorithm. **It was built on top of spacy for the exact purpose of phase extraction.** It borrows features from spacy and scores the phrases on the basis of Larry Page's PageRank algorithm. Hence, it makes perfect sense to use it here.\n", | |
| "\n", | |
| "You can find out more here : https://github.com/DerwenAI/pytextrank#kudos\n", | |
| "\n", | |
| "<img height=300 src=https://i.pinimg.com/originals/94/11/fc/9411fc3f50cb7c80734c4fa0ad8acf6a.jpg >" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "4XVF3ZbER_Lz" | |
| }, | |
| "source": [ | |
| "#-Create a sequential list of phrases as they appear in the text.\n", | |
| "\n", | |
| "import pytextrank\n", | |
| "\n", | |
| "keyphrases = []\n", | |
| "\n", | |
| "nlp = spacy.load(\"en_core_web_sm\")\n", | |
| "\n", | |
| "# add PyTextRank to the spaCy pipeline\n", | |
| "nlp.add_pipe(\"textrank\")\n", | |
| "\n", | |
| "for para in paragraphs : \n", | |
| " for sentence in para :\n", | |
| " doc = nlp(sentence)\n", | |
| "\n", | |
| " for p in doc._.phrases: # doc._.phrases is via the pipeline I added\n", | |
| " if len(p.text.split())>1: # to only extract phrases which have more than 1 word\n", | |
| " keyphrases.append(p.text) # saving the phrase" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "J1poruEUqgim", | |
| "outputId": "319d2a16-4409-43d9-808d-50021a851519" | |
| }, | |
| "source": [ | |
| "print(\"No. of keyphrases extracted : \",len(keyphrases),\"\\n\\n\")\n", | |
| "keyphrases" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "No. of keyphrases extracted : 9910 \n", | |
| "\n", | |
| "\n" | |
| ], | |
| "name": "stdout" | |
| }, | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/plain": [ | |
| "['MANY high school students',\n", | |
| " 'that earsplitting noise',\n", | |
| " 'a twisted perspective',\n", | |
| " 'a wailing monstrosity',\n", | |
| " 'a doubt',\n", | |
| " 'whose purpose',\n", | |
| " 'each morning',\n", | |
| " 'THE ALARM CLOCK',\n", | |
| " 'their eyes',\n", | |
| " 'the situation',\n", | |
| " 'hash browns',\n", | |
| " 'dear, sweet Sally',\n", | |
| " 'their kitchen tables',\n", | |
| " 'an alarm clock',\n", | |
| " 'a few extra hours',\n", | |
| " 'the air',\n", | |
| " 'that wafting, wondrous bacon smell',\n", | |
| " 'a Saturday morning',\n", | |
| " 'the chance',\n", | |
| " 'the taking',\n", | |
| " 'her slumber',\n", | |
| " 'another horribly boring day',\n", | |
| " 'A beautiful breakfast',\n", | |
| " 'the only thing',\n", | |
| " 'the scrumptious delight',\n", | |
| " 'these foolish Sallys',\n", | |
| " 'a slab',\n", | |
| " 'these kids',\n", | |
| " 'sheer waves',\n", | |
| " 'perhaps every school day',\n", | |
| " 'what a marvelous, glorious thing',\n", | |
| " 'those heart-stoppingly good strips',\n", | |
| " 'a plate',\n", | |
| " 'Oh, bacon',\n", | |
| " 'The thought',\n", | |
| " 'this lovely breakfast food',\n", | |
| " 'many others',\n", | |
| " 'my heart',\n", | |
| " 'a huge smile',\n", | |
| " 'my alarm clock',\n", | |
| " 'this same zeal',\n", | |
| " 'the mellifluous joy',\n", | |
| " 'the yearning',\n", | |
| " 'my face',\n", | |
| " 'every morning',\n", | |
| " 'new classes',\n", | |
| " 'Jay gatsby',\n", | |
| " 'diverse classes',\n", | |
| " 'AP Spanish Literature',\n", | |
| " 'new knowledge',\n", | |
| " 'good ol’ Yossarian',\n", | |
| " 'Spanish History',\n", | |
| " 'advanced students',\n", | |
| " 'a Middle Eastern History class',\n", | |
| " 'a new Spanish History lesson',\n", | |
| " 'my English Language class',\n", | |
| " 'any elusive bacon strip',\n", | |
| " 'a Middle Eastern History',\n", | |
| " 'any new bits',\n", | |
| " 'an intensive literature study',\n", | |
| " 'a faraway Magnet school',\n", | |
| " 'a strong attachment',\n", | |
| " 'my own pursuit',\n", | |
| " 'the post-AP exam period',\n", | |
| " 'their Magnet students',\n", | |
| " 'a specific aspect',\n", | |
| " 'my school bus',\n", | |
| " 'the “bacon mentality',\n", | |
| " 'the freeway',\n", | |
| " 'The Great Gatsby’s pursuit',\n", | |
| " 'Monday through Friday',\n", | |
| " 'a change',\n", | |
| " 'a subject',\n", | |
| " 'each day',\n", | |
| " 'the prospect',\n", | |
| " 'its benefits',\n", | |
| " 'a wealth',\n", | |
| " 'their thirty-minute rant',\n", | |
| " 'a classroom',\n", | |
| " 'even the day',\n", | |
| " 'The Great Gatsby’s',\n", | |
| " 'the opportunity',\n", | |
| " 'the enthusiasm',\n", | |
| " 'My peers',\n", | |
| " 'high school',\n", | |
| " 'fresh, new, crisp learning',\n", | |
| " 'some lackluster movie theatre',\n", | |
| " 'a renowned professor',\n", | |
| " 'another good book',\n", | |
| " 'my friends',\n", | |
| " 'all the new opportunities',\n", | |
| " 'all days',\n", | |
| " 'the day',\n", | |
| " 'the week',\n", | |
| " 'my time',\n", | |
| " 'all days of the week',\n", | |
| " 'my hunger',\n", | |
| " 'the weekend',\n", | |
| " 'a textbook',\n", | |
| " 'the university',\n", | |
| " 'another trip',\n", | |
| " 'my breakfast',\n", | |
| " 'six classes',\n", | |
| " 'all three meals',\n", | |
| " 'eager anticipation',\n", | |
| " 'US History',\n", | |
| " 'many factors',\n", | |
| " 'the day',\n", | |
| " 'A MATH-SCIENCE girl',\n", | |
| " 'the formulas',\n", | |
| " 'my fascination',\n", | |
| " 'later in the day',\n", | |
| " 'my success',\n", | |
| " 'other theorems',\n", | |
| " 'math justice',\n", | |
| " 'ancient Greece',\n", | |
| " 'a certain surface area',\n", | |
| " 'the grocery store',\n", | |
| " 'the real world',\n", | |
| " 'too little change',\n", | |
| " 'the Pythagorean Theorem',\n", | |
| " 'the store',\n", | |
| " 'Pascal’s triangle',\n", | |
| " 'the golden rectangle',\n", | |
| " 'a box',\n", | |
| " 'the box',\n", | |
| " 'the cumulative information',\n", | |
| " 'what math',\n", | |
| " 'the cashier',\n", | |
| " 'a host',\n", | |
| " 'the volume',\n", | |
| " 'the problems',\n", | |
| " 'One day',\n", | |
| " 'the fact',\n", | |
| " 'any person',\n", | |
| " 'nearly 12 years',\n", | |
| " 'all these once foreign symbols',\n", | |
| " 'its simplicity',\n", | |
| " 'the application',\n", | |
| " 'the calculations',\n", | |
| " 'The answer',\n", | |
| " 'The genius',\n", | |
| " 'these reasons',\n", | |
| " 'extra instruction',\n", | |
| " 'middle school',\n", | |
| " 'Calculus AB',\n", | |
| " 'a high or low speed math class',\n", | |
| " 'the slow speed math',\n", | |
| " 'review books',\n", | |
| " 'Incoming 6th graders',\n", | |
| " 'my junior year',\n", | |
| " 'the 5th grade',\n", | |
| " 'my classes',\n", | |
| " 'my grade',\n", | |
| " 'the second day',\n", | |
| " 'nearly a year of Calculus',\n", | |
| " 'a result',\n", | |
| " 'my older brother',\n", | |
| " 'a challenge',\n", | |
| " 'a C-',\n", | |
| " 'a B',\n", | |
| " 'nearly a year',\n", | |
| " 'a C',\n", | |
| " 'their scores',\n", | |
| " 'a lot',\n", | |
| " 'my parents',\n", | |
| " 'my time',\n", | |
| " 'a course',\n", | |
| " 'This persistence',\n", | |
| " 'My strongest subject',\n", | |
| " 'a test',\n", | |
| " 'an A.',\n", | |
| " 'a B+',\n", | |
| " 'my dedication',\n", | |
| " 'the summer',\n", | |
| " 'a transient love',\n", | |
| " 'a love',\n", | |
| " 'the love',\n", | |
| " 'the good times',\n", | |
| " 'FOUR YEARS OLD',\n", | |
| " 'a language',\n", | |
| " 'my side',\n", | |
| " 'a person',\n", | |
| " 'only language',\n", | |
| " 'Western influence',\n", | |
| " 'my innocent ear',\n", | |
| " 'its sweet and tender words',\n", | |
| " 'The only thing',\n", | |
| " 'the Iron Curtain',\n", | |
| " 'its perfectly articulated phrases',\n", | |
| " 'their opinion',\n", | |
| " 'a country',\n", | |
| " 'the matter',\n", | |
| " 'the truth',\n", | |
| " 'such a foreign',\n", | |
| " 'especially my family',\n", | |
| " 'Spanish language telenovelas',\n", | |
| " 'Romanian audiences',\n", | |
| " 'soap operas',\n", | |
| " 'new aspects',\n", | |
| " 'a Romanian cable network',\n", | |
| " 'her own language',\n", | |
| " 'June 16th',\n", | |
| " 'the local cable company',\n", | |
| " 'the language',\n", | |
| " 'the Romanian subtitles',\n", | |
| " 'the Spanish dialogue',\n", | |
| " 'the United States',\n", | |
| " 'our living room',\n", | |
| " 'a little girl',\n", | |
| " 'the world',\n", | |
| " 'the rest',\n", | |
| " 'no one',\n", | |
| " 'that nonsense',\n", | |
| " 'June 16th, 1994',\n", | |
| " 'a man',\n", | |
| " 'my apparent “obsession',\n", | |
| " 'quite an accomplishment',\n", | |
| " 'my door',\n", | |
| " 'your time',\n", | |
| " 'My father',\n", | |
| " 'little interaction',\n", | |
| " 'Los Angeles',\n", | |
| " 'many hardships',\n", | |
| " 'a huge and indispensable part',\n", | |
| " 'the language',\n", | |
| " 'a Spanish speaker’s paradise',\n", | |
| " 'my largely Romanian neighborhood',\n", | |
| " 'the feelings',\n", | |
| " 'the US',\n", | |
| " 'my family’s objection',\n", | |
| " 'my immigration',\n", | |
| " 'six years',\n", | |
| " 'Advanced Placement Spanish',\n", | |
| " 'Advanced Placement',\n", | |
| " 'high school',\n", | |
| " 'a new era',\n", | |
| " 'the language',\n", | |
| " 'an era',\n", | |
| " 'my life',\n", | |
| " 'my love',\n", | |
| " 'an hour',\n", | |
| " 'two years',\n", | |
| " 'a great asset',\n", | |
| " 'the greatest compliment',\n", | |
| " 'the second most spoken language',\n", | |
| " 'the language',\n", | |
| " 'my Romanian last name',\n", | |
| " 'their objections',\n", | |
| " 'whose family members',\n", | |
| " 'the natives',\n", | |
| " 'a Romanian girl',\n", | |
| " 'my ability',\n", | |
| " 'my husband',\n", | |
| " 'a recent trip',\n", | |
| " 'the United States',\n", | |
| " 'my father',\n", | |
| " 'One man',\n", | |
| " 'no words',\n", | |
| " 'unstoppable perseverance',\n", | |
| " 'family objections',\n", | |
| " 'whose native language',\n", | |
| " 'the primary language',\n", | |
| " 'a professional and personal level',\n", | |
| " 'an ambitious and aggressive spirit',\n", | |
| " 'the United States',\n", | |
| " '21 nations',\n", | |
| " 'these nations',\n", | |
| " 'the six official languages',\n", | |
| " 'a new land',\n", | |
| " 'the face',\n", | |
| " 'the Un',\n", | |
| " 'the world',\n", | |
| " 'the globe',\n", | |
| " 'a part',\n", | |
| " 'the people',\n", | |
| " 'the link',\n", | |
| " 'an impact',\n", | |
| " 'My love',\n", | |
| " 'my love',\n", | |
| " 'my ability',\n", | |
| " 'the 40 million Americans',\n", | |
| " 'The fight',\n", | |
| " 'the path',\n", | |
| " '40 million',\n", | |
| " 'such passion',\n", | |
| " 'Los Angeles',\n", | |
| " 'both her native and adoptive countries',\n", | |
| " 'a big part',\n", | |
| " 'a language',\n", | |
| " 'an even bigger part',\n", | |
| " 'my most remarkable accomplishments',\n", | |
| " 'a Romanian-American girl',\n", | |
| " 'the future',\n", | |
| " 'the rest',\n", | |
| " 'the case',\n", | |
| " 'the years',\n", | |
| " 'Teenage Female Angst',\n", | |
| " 'valuable information',\n", | |
| " 'Holden Caulfield',\n", | |
| " 'everyday occurrences',\n", | |
| " 'seemingly mindless and trashy zombie films',\n", | |
| " 'Clean Thoughts',\n", | |
| " 'sexual imagery',\n", | |
| " 'profane or vulgar language',\n", | |
| " 'Honors Program',\n", | |
| " 'Dirty Words',\n", | |
| " 'the Vampire Slayer',\n", | |
| " 'the few subtle and transcendent moments',\n", | |
| " 'Madame Bovary',\n", | |
| " 'a cracked kettle',\n", | |
| " 'the only course',\n", | |
| " 'the best and brightest students',\n", | |
| " 'a six-week intensive college-like experience',\n", | |
| " 'the best thing',\n", | |
| " 'the stars',\n", | |
| " 'the exercise',\n", | |
| " 'THE governor’s Honors Program',\n", | |
| " 'a brand-new way',\n", | |
| " 'The classes',\n", | |
| " 'a class',\n", | |
| " 'the classes',\n", | |
| " 'The experiences',\n", | |
| " 'the “classics',\n", | |
| " 'the bears',\n", | |
| " 'the use',\n", | |
| " 'the varying education',\n", | |
| " 'the world',\n", | |
| " 'a hackneyed phrase',\n", | |
| " 'the extremes',\n", | |
| " 'seventeen year-olds',\n", | |
| " 'the too-close confines',\n", | |
| " 'my mind',\n", | |
| " 'sixteen and seventeen year-olds',\n", | |
| " 'THIS SUMMER',\n", | |
| " 'the teachers',\n", | |
| " 'the only thing',\n", | |
| " 'the things',\n", | |
| " 'the classroom',\n", | |
| " 'the classrooms',\n", | |
| " 'the people',\n", | |
| " 'an interesting perspective',\n", | |
| " 'the world',\n", | |
| " 'a big name',\n", | |
| " 'My best friends',\n", | |
| " 'the running joke',\n", | |
| " 'my growth',\n", | |
| " 'the ethical framework',\n", | |
| " 'an accomplished singer',\n", | |
| " 'an endearing trick',\n", | |
| " 'their beliefs',\n", | |
| " 'the former’s girl-friend',\n", | |
| " 'an excellent athlete',\n", | |
| " 'a phenomenal writer',\n", | |
| " 'my clashes',\n", | |
| " 'a math major',\n", | |
| " 'fairly conservative Christians',\n", | |
| " 'every sport',\n", | |
| " 'the sort',\n", | |
| " 'a lot',\n", | |
| " 'The experience',\n", | |
| " 'college undergrads',\n", | |
| " 'Minority Introduction',\n", | |
| " '68 other rising seniors',\n", | |
| " 'a highly rigorous academic program',\n", | |
| " 'a small auditorium',\n", | |
| " 'the program',\n", | |
| " 'our lives',\n", | |
| " 'THIS PAST SUMMER',\n", | |
| " 'our classes',\n", | |
| " 'the top',\n", | |
| " 'THE opportunity',\n", | |
| " 'our minds',\n", | |
| " 'our confidence',\n", | |
| " 'six and a half weeks',\n", | |
| " 'hard times',\n", | |
| " 'problem sets',\n", | |
| " 'drill motors',\n", | |
| " 'my best friends',\n", | |
| " 'the same time',\n", | |
| " 'MITES valedictorians',\n", | |
| " 'my support system',\n", | |
| " 'a semiautonomous robot',\n", | |
| " 'even my family',\n", | |
| " 'these strangers',\n", | |
| " 'the second week',\n", | |
| " 'each other’s shoulders',\n", | |
| " 'the morning',\n", | |
| " 'a book per week',\n", | |
| " 'a book',\n", | |
| " 'The 16 hours days',\n", | |
| " 'four hours',\n", | |
| " 'two in the morning',\n", | |
| " 'a well known fact',\n", | |
| " 'an average',\n", | |
| " 'AP Physics',\n", | |
| " 'introductory level engineering',\n", | |
| " 'first year',\n", | |
| " 'AP literature',\n", | |
| " 'second year',\n", | |
| " 'Introductory Physics',\n", | |
| " 'Engineering Design',\n", | |
| " 'Introductory Physics, Engineering Design',\n", | |
| " 'Physics and Chemistry',\n", | |
| " 'a full year',\n", | |
| " 'no physics experience',\n", | |
| " 'the only A',\n", | |
| " 'a semester',\n", | |
| " 'the material',\n", | |
| " 'the fall',\n", | |
| " 'my schedule',\n", | |
| " '5 classes',\n", | |
| " 'a half',\n", | |
| " 'The work',\n", | |
| " 'the month',\n", | |
| " 'the month and a half',\n", | |
| " 'folklorico dancing',\n", | |
| " 'Science Bowl team',\n", | |
| " 'swim team',\n", | |
| " 'ACE mentoring team',\n", | |
| " 'other teams',\n", | |
| " 'educational opportunities',\n", | |
| " 'Jewish Student Union',\n", | |
| " 'California Honors Society',\n", | |
| " 'Play Production',\n", | |
| " 'Scholarship Federation',\n", | |
| " 'Science Bowl',\n", | |
| " 'GEAR-UP Mentoring Program',\n", | |
| " 'California Honors Society and Scholarship Federation',\n", | |
| " 'new programs',\n", | |
| " 'This academic school year',\n", | |
| " 'the Speech and Debate team',\n", | |
| " 'my school',\n", | |
| " 'the regular coursework',\n", | |
| " 'the peak',\n", | |
| " 'my involvement',\n", | |
| " 'this program',\n", | |
| " 'the lookout',\n", | |
| " 'my academic endeavors',\n", | |
| " 'my academic pursuits',\n", | |
| " 'the bar',\n", | |
| " 'my life',\n", | |
| " 'the single most pivotal point',\n", | |
| " 'the most challenging experience',\n", | |
| " 'The program',\n", | |
| " 'the program',\n", | |
| " 'The assistants',\n", | |
| " 'The motivation',\n", | |
| " 'my breath',\n", | |
| " 'My eyes',\n", | |
| " 'The Apology',\n", | |
| " 'incoherent words',\n", | |
| " 'the adjacent words',\n", | |
| " 'a simple word',\n", | |
| " 'a familiar word',\n", | |
| " 'the last word',\n", | |
| " 'the sixth word',\n", | |
| " 'a word',\n", | |
| " 'one word',\n", | |
| " 'an incomplete translation',\n", | |
| " 'a coherent translation',\n", | |
| " 'the most important word',\n", | |
| " 'a deep breath',\n", | |
| " 'The Apology',\n", | |
| " 'the clause',\n", | |
| " 'my senses',\n", | |
| " 'the mercy',\n", | |
| " '20 lines',\n", | |
| " 'My initial scan',\n", | |
| " 'The Greek letters',\n", | |
| " 'the bell rings',\n", | |
| " 'the text',\n", | |
| " 'the page',\n", | |
| " 'my head',\n", | |
| " 'the lexicon',\n", | |
| " 'My eyes',\n", | |
| " 'The whirlwind',\n", | |
| " 'figure skating',\n", | |
| " 'different mediums',\n", | |
| " 'hard work',\n", | |
| " 'the value',\n", | |
| " 'The Apology',\n", | |
| " 'cold wind',\n", | |
| " 'thin sheets',\n", | |
| " 'a new bruise',\n", | |
| " 'My body',\n", | |
| " 'my body',\n", | |
| " 'the music',\n", | |
| " 'the next jump',\n", | |
| " 'my eyes',\n", | |
| " 'a sharp skid',\n", | |
| " 'my next program',\n", | |
| " 'the air',\n", | |
| " 'a visual performance',\n", | |
| " 'three times',\n", | |
| " 'a bright January morning',\n", | |
| " 'a jump',\n", | |
| " 'the ice',\n", | |
| " 'my knees',\n", | |
| " 'my face',\n", | |
| " 'the translation',\n", | |
| " 'my torso',\n", | |
| " 'my arms',\n", | |
| " 'every note',\n", | |
| " 'every pitch',\n", | |
| " 'the cadenza',\n", | |
| " 'the ground',\n", | |
| " 'A shock',\n", | |
| " 'the midst',\n", | |
| " 'my frame',\n", | |
| " 'Meadow Lake',\n", | |
| " 'the proper removal process',\n", | |
| " 'a Mugwort removal cleanup',\n", | |
| " 'a strenuous month',\n", | |
| " 'The tedious logistical process',\n", | |
| " 'a Discovery Guide Leader',\n", | |
| " 'July 2018',\n", | |
| " 'a permit',\n", | |
| " 'the importance',\n", | |
| " 'my activism',\n", | |
| " 'my plan',\n", | |
| " 'the proper removal technique',\n", | |
| " 'my eager group',\n", | |
| " 'an invasive species',\n", | |
| " 'a productive plan',\n", | |
| " 'the shore',\n", | |
| " 'that knowledge',\n", | |
| " 'the soil',\n", | |
| " 'the stem',\n", | |
| " 'each person',\n", | |
| " 'the shovel',\n", | |
| " 'The hot sun',\n", | |
| " 'my back',\n", | |
| " 'the plant',\n", | |
| " '4.2 pounds',\n", | |
| " 'the benefits',\n", | |
| " 'its hold',\n", | |
| " 'The ground',\n", | |
| " 'Twenty pairs',\n", | |
| " 'the day',\n", | |
| " 'more plans',\n", | |
| " 'more songs',\n", | |
| " 'many things',\n", | |
| " 'my actions',\n", | |
| " 'Plato years ago',\n", | |
| " 'my knowledge',\n", | |
| " 'the lessons',\n", | |
| " 'the world',\n", | |
| " 'The Apology',\n", | |
| " 'the math packet',\n", | |
| " 'a mental contract',\n", | |
| " 'your pages',\n", | |
| " 'the packet',\n", | |
| " 'the black italicized letters',\n", | |
| " 'my life',\n", | |
| " 'my arms',\n", | |
| " 'the top bunk',\n", | |
| " 'a challenge',\n", | |
| " 'a pencil',\n", | |
| " 'my hand',\n", | |
| " 'the title',\n", | |
| " 'all the problems',\n", | |
| " 'the goosebumps',\n", | |
| " 'the mysteries',\n", | |
| " 'her words',\n", | |
| " 'my teacher',\n", | |
| " 'my fingers',\n", | |
| " 'the cover',\n", | |
| " 'better pay',\n", | |
| " 'her minimum wage job',\n", | |
| " 'a makeshift bed',\n", | |
| " 'the second bed',\n", | |
| " 'a torn and scratchy old sofa',\n", | |
| " 'a bed',\n", | |
| " 'one bed',\n", | |
| " 'the creative one',\n", | |
| " 'our new home',\n", | |
| " 'one job',\n", | |
| " 'that little room',\n", | |
| " 'my own experiences',\n", | |
| " 'the omnipresent challenge',\n", | |
| " 'the many challenges',\n", | |
| " 'My dad',\n", | |
| " 'my dad',\n", | |
| " 'My mom',\n", | |
| " 'my mom',\n", | |
| " 'his pursuit',\n", | |
| " 'another challenge',\n", | |
| " 'their hard work',\n", | |
| " 'a bunk',\n", | |
| " 'my family',\n", | |
| " 'the Atlantic',\n", | |
| " 'my life',\n", | |
| " 'late working',\n", | |
| " 'our first year',\n", | |
| " 'her crib',\n", | |
| " 'our belongings',\n", | |
| " 'the comfort',\n", | |
| " 'the tenacity',\n", | |
| " 'the consequences',\n", | |
| " 'My love',\n", | |
| " 'My parents',\n", | |
| " 'My sister',\n", | |
| " 'each day',\n", | |
| " 'one example',\n", | |
| " 'Ten years ago',\n", | |
| " 'just another stage',\n", | |
| " 'the value',\n", | |
| " 'these values',\n", | |
| " 'a single job',\n", | |
| " 'my dad’s consistent job switches',\n", | |
| " 'my mom’s devotion',\n", | |
| " 'my family',\n", | |
| " 'my sisters',\n", | |
| " 'any challenge',\n", | |
| " 'the packet',\n", | |
| " 'a grueling two months',\n", | |
| " 'two months',\n", | |
| " 'my own experiences',\n", | |
| " 'all the questions',\n", | |
| " 'my parents',\n", | |
| " 'My 11-year-old self',\n", | |
| " 'our ten years',\n", | |
| " 'middle and high school',\n", | |
| " 'These learned values',\n", | |
| " 'my school',\n", | |
| " 'my goals',\n", | |
| " 'this goal',\n", | |
| " 'my value',\n", | |
| " 'these values',\n", | |
| " 'the most challenging courses',\n", | |
| " 'the past five years',\n", | |
| " 'the guidance',\n", | |
| " 'the people',\n", | |
| " 'a surgeon',\n", | |
| " 'a career',\n", | |
| " 'my church',\n", | |
| " 'the person',\n", | |
| " 'the way',\n", | |
| " 'my community',\n", | |
| " 'my time',\n", | |
| " 'muddy soccer cleats',\n", | |
| " 'high grades',\n", | |
| " 'the same successes',\n", | |
| " 'the same results',\n", | |
| " 'my field hockey gear',\n", | |
| " '15 field hockey team',\n", | |
| " 'the wooden bench',\n", | |
| " 'my cleats',\n", | |
| " 'my locker',\n", | |
| " 'my heavy legs',\n", | |
| " 'a wet and dreary October evening',\n", | |
| " 'the concrete',\n", | |
| " 'the Netherlands',\n", | |
| " 'my homework',\n", | |
| " 'the year',\n", | |
| " 'that practice',\n", | |
| " 'my head',\n", | |
| " 'the dirt',\n", | |
| " 'The sound',\n", | |
| " 'field hockey',\n", | |
| " 'field hockey cleats',\n", | |
| " 'new opportunities',\n", | |
| " 'Deerfield’s field hockey coach',\n", | |
| " 'gender boundaries',\n", | |
| " 'a girls’ field hockey practice',\n", | |
| " 'my tough transition',\n", | |
| " 'more eyes',\n", | |
| " 'a warm rush',\n", | |
| " 'Ms. McVaugh',\n", | |
| " 'the unusual opportunity',\n", | |
| " 'our red faces',\n", | |
| " 'the gender barrier',\n", | |
| " 'the shared fanaticism',\n", | |
| " 'the bench',\n", | |
| " 'our clamor',\n", | |
| " 'a burning glow',\n", | |
| " 'my initial anxiety',\n", | |
| " 'my exuberance',\n", | |
| " 'the girls',\n", | |
| " 'my veins',\n", | |
| " 'the following day',\n", | |
| " 'my breath',\n", | |
| " 'the emotion',\n", | |
| " 'each second',\n", | |
| " 'my body',\n", | |
| " 'the thought',\n", | |
| " 'the warm-up',\n", | |
| " 'each step',\n", | |
| " 'the ball',\n", | |
| " 'the turf',\n", | |
| " 'an answer',\n", | |
| " 'that moment',\n", | |
| " 'the end',\n", | |
| " 'the whistle',\n", | |
| " 'new skills',\n", | |
| " 'its new athletic challenges',\n", | |
| " 'my new approach',\n", | |
| " 'the soccer field',\n", | |
| " 'each new improvement',\n", | |
| " 'the sport',\n", | |
| " 'the same atmosphere',\n", | |
| " 'the value',\n", | |
| " 'this experience',\n", | |
| " 'the joy',\n", | |
| " 'the turf',\n", | |
| " 'my focus',\n", | |
| " 'such successes',\n", | |
| " 'such moments',\n", | |
| " 'new material',\n", | |
| " 'boarding school',\n", | |
| " 'my field hockey gear',\n", | |
| " 'organic chemistry',\n", | |
| " 'my new mindset',\n", | |
| " 'deeper questions',\n", | |
| " 'even athletic achievements',\n", | |
| " 'the other opportunities',\n", | |
| " 'Dutch ‘stroopwafels',\n", | |
| " 'the same open, curious, and risk-taking mindset',\n", | |
| " 'nucleophile-electrophile reactions',\n", | |
| " 'the exploration',\n", | |
| " 'the grades',\n", | |
| " 'the cultural differences',\n", | |
| " 'my hall',\n", | |
| " 'my peers',\n", | |
| " 'a result',\n", | |
| " 'the classroom',\n", | |
| " 'the dorm',\n", | |
| " 'such opportunities',\n", | |
| " 'their experiences',\n", | |
| " 'the obstacles',\n", | |
| " 'the successes',\n", | |
| " 'my dorm',\n", | |
| " 'the classroom',\n", | |
| " 'my cleats',\n", | |
| " 'my heart',\n", | |
| " 'a lingering poor taste',\n", | |
| " 'a creamy white center',\n", | |
| " 'my mouth',\n", | |
| " 'a firm chocolate crust',\n", | |
| " 'an Oreo',\n", | |
| " 'a social meaning',\n", | |
| " 'the exterior',\n", | |
| " 'confused stares',\n", | |
| " 'a new story',\n", | |
| " 'Hans Zimmer’s',\n", | |
| " 'both Hans Zimmer',\n", | |
| " 'high school',\n", | |
| " 'Top 40 hits',\n", | |
| " 'each note',\n", | |
| " 'a little girl',\n", | |
| " 'my black peers',\n", | |
| " 'the upbeat, peppy nature',\n", | |
| " 'a tango melody',\n", | |
| " 'my musical taste',\n", | |
| " 'vastly different styles',\n", | |
| " 'Ballroom and pop music',\n", | |
| " 'the seductive, powerful attacks',\n", | |
| " 'the whitest thing',\n", | |
| " 'a friend',\n", | |
| " 'such an Oreo',\n", | |
| " 'new guilty sensations',\n", | |
| " 'white people',\n", | |
| " 'social alienation',\n", | |
| " 'their black roots',\n", | |
| " 'a black person',\n", | |
| " 'my skin tone',\n", | |
| " 'an African-American woman',\n", | |
| " 'a certain race',\n", | |
| " 'ballroom and pop music',\n", | |
| " 'the negatively charged implications',\n", | |
| " 'my so-called white core',\n", | |
| " 'my musical interests',\n", | |
| " 'the expectations',\n", | |
| " 'an Oreo',\n", | |
| " 'two seemingly different commodities',\n", | |
| " 'the connection',\n", | |
| " 'various musical styles',\n", | |
| " 'the county solo & ensemble festival',\n", | |
| " 'Lyrical Composition',\n", | |
| " 'each day',\n", | |
| " 'my band director',\n", | |
| " 'the page',\n", | |
| " 'a solo',\n", | |
| " 'a clarinet player',\n", | |
| " 'the rhythms',\n", | |
| " 'my band',\n", | |
| " 'my freshman year',\n", | |
| " 'my life',\n", | |
| " 'the notes',\n", | |
| " 'a piece',\n", | |
| " 'the right time',\n", | |
| " 'the right note',\n", | |
| " 'this piece',\n", | |
| " 'your interpretation',\n", | |
| " 'other musical markings',\n", | |
| " 'breath marks',\n", | |
| " 'black ink',\n", | |
| " 'white space',\n", | |
| " 'a climactic precipice',\n", | |
| " 'my own contribution',\n", | |
| " 'first glance',\n", | |
| " 'second glance',\n", | |
| " 'the black ink',\n", | |
| " 'a dynamic journey',\n", | |
| " 'such a definitive nature',\n", | |
| " 'the measures',\n", | |
| " 'my pencil',\n", | |
| " 'the sheet',\n", | |
| " 'a composition',\n", | |
| " 'this composition',\n", | |
| " 'the audience',\n", | |
| " 'the distinction',\n", | |
| " 'the rhythms',\n", | |
| " 'This realization',\n", | |
| " 'the notes',\n", | |
| " 'certain races',\n", | |
| " 'Most stereotypes',\n", | |
| " 'the documented definition',\n", | |
| " 'the mind',\n", | |
| " 'the minds',\n", | |
| " 'an Oreo',\n", | |
| " 'the black woman society',\n", | |
| " 'a proud black woman',\n", | |
| " 'ballroom music',\n", | |
| " 'the black experience',\n", | |
| " 'my own music maker',\n", | |
| " 'my own input',\n", | |
| " 'a beautiful addition',\n", | |
| " 'my own design',\n", | |
| " 'a blank canvas',\n", | |
| " 'The creamy white center',\n", | |
| " 'my ever-evolving character',\n", | |
| " 'my mark',\n", | |
| " 'a One Direction tune',\n", | |
| " 'a betrayal',\n", | |
| " 'My musical interests',\n", | |
| " 'my definition',\n", | |
| " 'my desire',\n", | |
| " 'my roots',\n", | |
| " 'its grip',\n", | |
| " 'the intricacies',\n", | |
| " 'an Oreo',\n", | |
| " 'my choice',\n", | |
| " 'my style',\n", | |
| " 'my parents',\n", | |
| " 'the clock',\n", | |
| " 'a letter',\n", | |
| " 'a sibling',\n", | |
| " 'my friends',\n", | |
| " 'Lhasa Apso dog',\n", | |
| " 'an only child',\n", | |
| " 'my new sister',\n", | |
| " 'a whole',\n", | |
| " 'the other ways',\n", | |
| " 'a dog',\n", | |
| " 'my life',\n", | |
| " 'my community',\n", | |
| " 'my group',\n", | |
| " 'my parents',\n", | |
| " 'a furry',\n", | |
| " 'these sentiments',\n", | |
| " 'world history classes',\n", | |
| " 'East Asian',\n", | |
| " 'vast differences',\n", | |
| " 'less importance',\n", | |
| " 'social studies classes',\n", | |
| " 'European and East Asian history',\n", | |
| " 'cheese sandwiches',\n", | |
| " 'the only other Filipino girl',\n", | |
| " 'the only way',\n", | |
| " 'the traditional adobo chicken',\n", | |
| " 'my immigrant parents',\n", | |
| " 'the only pathway',\n", | |
| " 'my Filipino self',\n", | |
| " 'the Filipino delicacies',\n", | |
| " 'a huge impact',\n", | |
| " 'our cultures',\n", | |
| " 'my friends',\n", | |
| " 'about 75% white students',\n", | |
| " 'my community',\n", | |
| " 'the way',\n", | |
| " 'the norm',\n", | |
| " 'my grade',\n", | |
| " 'my mom',\n", | |
| " 'the ancestry',\n", | |
| " 'one category',\n", | |
| " 'a predominantly white town',\n", | |
| " 'the pungent aromas',\n", | |
| " 'a population',\n", | |
| " 'a school',\n", | |
| " 'my heritage',\n", | |
| " 'about 75%',\n", | |
| " 'the notion',\n", | |
| " 'my classmates',\n", | |
| " 'different viewpoints',\n", | |
| " 'Young Conservatives Club',\n", | |
| " 'current state',\n", | |
| " 'high school',\n", | |
| " 'the High School Democrats Club and Young Conservatives Club',\n", | |
| " 'the High School Democrats Club',\n", | |
| " 'Filipino dishes',\n", | |
| " 'my Filipino pride',\n", | |
| " 'my social studies',\n", | |
| " 'my freshman year',\n", | |
| " 'the student body',\n", | |
| " 'my feelings',\n", | |
| " 'my school',\n", | |
| " 'the Philippines',\n", | |
| " 'speech and debate tournaments',\n", | |
| " 'my background',\n", | |
| " 'the history',\n", | |
| " 'an increase',\n", | |
| " 'the spectrum',\n", | |
| " 'the globe',\n", | |
| " 'an array',\n", | |
| " 'my friends',\n", | |
| " 'my peers',\n", | |
| " 'the message',\n", | |
| " 'my Filipino background',\n", | |
| " 'my appreciation',\n", | |
| " 'the norm',\n", | |
| " 'the potential',\n", | |
| " 'the community',\n", | |
| " 'my narrative',\n", | |
| " 'most conventional classrooms',\n", | |
| " 'the following day',\n", | |
| " 'this past summer',\n", | |
| " 'a test',\n", | |
| " 'other things',\n", | |
| " 'paper planes',\n", | |
| " 'New Jersey',\n", | |
| " 'more water',\n", | |
| " 'solar ovens',\n", | |
| " 'light and space vacuums',\n", | |
| " 'best and brightest students',\n", | |
| " 'Physical Science',\n", | |
| " '200 students',\n", | |
| " 'New Jersey’s',\n", | |
| " 'a topic',\n", | |
| " 'the speed',\n", | |
| " 'the content',\n", | |
| " 'the next weeks',\n", | |
| " 'the dynamic',\n", | |
| " 'a college campus',\n", | |
| " 'the SPK Program',\n", | |
| " 'a box',\n", | |
| " 'the box',\n", | |
| " 'a five-week enrichment program',\n", | |
| " 'the top',\n", | |
| " 'no textbooks',\n", | |
| " 'the table',\n", | |
| " 'the first day',\n", | |
| " 'a siphon',\n", | |
| " 'no phones',\n", | |
| " 'our teacher',\n", | |
| " 'just our brains',\n", | |
| " 'impressive terminology',\n", | |
| " 'the AP Physics exam',\n", | |
| " 'the AP Physics',\n", | |
| " 'a better future',\n", | |
| " 'our community',\n", | |
| " 'a competition',\n", | |
| " 'the problems',\n", | |
| " 'the means',\n", | |
| " 'those five weeks',\n", | |
| " 'many tropical diseases',\n", | |
| " 'Yellow fever',\n", | |
| " 'yellow fever',\n", | |
| " 'blood tests',\n", | |
| " 'several hours',\n", | |
| " 'an urban hospital',\n", | |
| " 'My skin',\n", | |
| " 'the local doctor',\n", | |
| " 'the time',\n", | |
| " 'a diagnosis',\n", | |
| " 'such a solvable issue',\n", | |
| " 'my family',\n", | |
| " 'small changes',\n", | |
| " 'neighboring villages',\n", | |
| " 'health education programs',\n", | |
| " 'most other organizations',\n", | |
| " 'immense impacts',\n", | |
| " 'several universities',\n", | |
| " 'small stands',\n", | |
| " 'simple things',\n", | |
| " 'real life problems',\n", | |
| " 'medical clinics',\n", | |
| " 'water sanitation plants',\n", | |
| " 'third parties',\n", | |
| " 'branch president',\n", | |
| " 'public gatherings',\n", | |
| " 'a non-profit organization',\n", | |
| " 'health and social issues',\n", | |
| " 'the small village',\n", | |
| " 'the lifestyle changes',\n", | |
| " 'a big effect',\n", | |
| " 'its basic needs',\n", | |
| " 'the first high school branch',\n", | |
| " 'one village',\n", | |
| " ...]" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 17 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "q_P2oNJTYnyr" | |
| }, | |
| "source": [ | |
| "**BELOW IS ANOTHER EXAMPLE OF HOW PYTEXTRANK WORKS AND HOW WELL IT PERFORMS**" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 332 | |
| }, | |
| "id": "HvWg72dWkV4d", | |
| "outputId": "94d1acfa-60f6-4e42-e421-f0e3f9bbb03a" | |
| }, | |
| "source": [ | |
| "import timeit\n", | |
| "import pytextrank\n", | |
| "y = []\n", | |
| "keywords = []\n", | |
| "text_data = \"Trump Impeachment Trial : The Senate Judiciary Committee will hold a confirmation hearing in two weeks for President Joe Biden’s pick to lead the Justice Department, Merrick Garland. The hearing, scheduled for Feb. 22 and 23, sets Garland up for a March 1 vote out of committee and comes after Senate Judiciary Committee Chair Dick Durbin (D-Ill.) and Ranking Member Chuck Grassley (R-Iowa) reached an agreement for the schedule.\"\n", | |
| "\n", | |
| "start = timeit.default_timer()\n", | |
| "nlp = spacy.load(\"en_core_web_sm\") #loading spacy\n", | |
| "stop = timeit.default_timer()\n", | |
| "print('Time to load spaCy model: ', stop - start)\n", | |
| "y.append(stop - start)\n", | |
| "\n", | |
| "nlp.add_pipe(\"textrank\", last=True) #adding pipeline \n", | |
| "\n", | |
| "start = timeit.default_timer()\n", | |
| "doc = nlp(text_data) #running on the text\n", | |
| "stop = timeit.default_timer()\n", | |
| "print('Time to run NLP on text: ', stop - start)\n", | |
| "y.append(stop - start)\n", | |
| "\n", | |
| "start = timeit.default_timer()\n", | |
| "for p in doc._.phrases:\n", | |
| " if p.rank>0.05 and len(p.text.split())<=4: # p.rank is relevancy score for that phrase\n", | |
| " k=[]\n", | |
| " k.append(p.text) #keyword\n", | |
| " k.append(float(\"%.6f\"%p.rank)) #score\n", | |
| " k.append(p.count) #frequency\n", | |
| " keywords.append(k)\n", | |
| "stop = timeit.default_timer()\n", | |
| "print('Time to extract and store keywords: ', stop - start)\n", | |
| "y.append(stop - start)\n", | |
| "\n", | |
| "x=['spaCy', 'NLP on text', 'keywords']\n", | |
| "import seaborn as sns\n", | |
| "plot = sns.barplot(x=x,y=y,hue=x).set_title(\"TIME TAKEN TO LOAD COMPONENTS\") " | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Time to load spaCy model: 0.4118130800000017\n", | |
| "Time to run NLP on text: 0.035720359000009694\n", | |
| "Time to extract and store keywords: 0.0001724580000086462\n" | |
| ], | |
| "name": "stdout" | |
| }, | |
| { | |
| "output_type": "display_data", | |
| "data": { | |
| "image/png": "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\n", | |
| "text/plain": [ | |
| "<Figure size 432x288 with 1 Axes>" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [], | |
| "needs_background": "light" | |
| } | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "BMbW24z4f2ii" | |
| }, | |
| "source": [ | |
| "**STEP 4** - N-gram next phrase prediction model using either ML/DL" | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "ZXD_B6_5Y2bl" | |
| }, | |
| "source": [ | |
| "# WORDCLOUD\n", | |
| "A wordcloud can be useful to see the pre-dominant words that are present in our corpus before going in to build the N-gram prediction model" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/", | |
| "height": 364 | |
| }, | |
| "id": "RnWWJ3rPuY1P", | |
| "outputId": "67006702-031b-4920-864e-31a159ddba74" | |
| }, | |
| "source": [ | |
| "import matplotlib.pyplot as plt\n", | |
| "from wordcloud import WordCloud\n", | |
| "\n", | |
| "plt.figure(figsize = (10,10))\n", | |
| "ax = plt.gca()\n", | |
| "ax.axes.xaxis.set_visible(False) # hiding axes\n", | |
| "ax.axes.yaxis.set_visible(False)\n", | |
| "wc = WordCloud(max_words = 100 , width = 1600 , height = 900, background_color='white',scale=2.5).generate(\" \".join(keyphrases))\n", | |
| "plt.imshow(wc)\n", | |
| "plt.title(\"WORDCLOUD\")\n", | |
| "plt.show()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "display_data", | |
| "data": { | |
| "image/png": "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\n", | |
| "text/plain": [ | |
| "<Figure size 720x720 with 1 Axes>" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [], | |
| "needs_background": "light" | |
| } | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "IlPm1zTTf417", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "outputId": "313c7b12-3036-4c8f-ec2d-0731b003b86a" | |
| }, | |
| "source": [ | |
| "#Implement N-gram next phrase prediction model.\n", | |
| "\n", | |
| "text=[ preprocess(i) for i in keyphrases ]\n", | |
| "text" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "[nltk_data] Downloading package stopwords to /root/nltk_data...\n", | |
| "[nltk_data] Unzipping corpora/stopwords.zip.\n" | |
| ], | |
| "name": "stdout" | |
| }, | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/plain": [ | |
| "['many high school students',\n", | |
| " 'earsplitting noise',\n", | |
| " 'twisted perspective',\n", | |
| " 'wailing monstrosity',\n", | |
| " 'doubt',\n", | |
| " 'whose purpose',\n", | |
| " 'morning',\n", | |
| " 'alarm clock',\n", | |
| " 'eyes',\n", | |
| " 'situation',\n", | |
| " 'hash browns',\n", | |
| " 'dear sweet sally',\n", | |
| " 'kitchen tables',\n", | |
| " 'alarm clock',\n", | |
| " 'extra hours',\n", | |
| " 'air',\n", | |
| " 'wafting wondrous bacon smell',\n", | |
| " 'saturday morning',\n", | |
| " 'chance',\n", | |
| " 'taking',\n", | |
| " 'slumber',\n", | |
| " 'another horribly boring day',\n", | |
| " 'beautiful breakfast',\n", | |
| " 'thing',\n", | |
| " 'scrumptious delight',\n", | |
| " 'foolish sallys',\n", | |
| " 'slab',\n", | |
| " 'kids',\n", | |
| " 'sheer waves',\n", | |
| " 'perhaps every school day',\n", | |
| " 'marvelous glorious thing',\n", | |
| " 'heart stoppingly good strips',\n", | |
| " 'plate',\n", | |
| " 'oh bacon',\n", | |
| " 'thought',\n", | |
| " 'lovely breakfast food',\n", | |
| " 'many others',\n", | |
| " 'heart',\n", | |
| " 'huge smile',\n", | |
| " 'alarm clock',\n", | |
| " 'zeal',\n", | |
| " 'mellifluous joy',\n", | |
| " 'yearning',\n", | |
| " 'face',\n", | |
| " 'every morning',\n", | |
| " 'new classes',\n", | |
| " 'jay gatsby',\n", | |
| " 'diverse classes',\n", | |
| " 'ap spanish literature',\n", | |
| " 'new knowledge',\n", | |
| " 'good ol yossarian',\n", | |
| " 'spanish history',\n", | |
| " 'advanced students',\n", | |
| " 'middle eastern history class',\n", | |
| " 'new spanish history lesson',\n", | |
| " 'english language class',\n", | |
| " 'elusive bacon strip',\n", | |
| " 'middle eastern history',\n", | |
| " 'new bits',\n", | |
| " 'intensive literature study',\n", | |
| " 'faraway magnet school',\n", | |
| " 'strong attachment',\n", | |
| " 'pursuit',\n", | |
| " 'post ap exam period',\n", | |
| " 'magnet students',\n", | |
| " 'specific aspect',\n", | |
| " 'school bus',\n", | |
| " 'bacon mentality',\n", | |
| " 'freeway',\n", | |
| " 'great gatsby pursuit',\n", | |
| " 'monday friday',\n", | |
| " 'change',\n", | |
| " 'subject',\n", | |
| " 'day',\n", | |
| " 'prospect',\n", | |
| " 'benefits',\n", | |
| " 'wealth',\n", | |
| " 'thirty minute rant',\n", | |
| " 'classroom',\n", | |
| " 'even day',\n", | |
| " 'great gatsby',\n", | |
| " 'opportunity',\n", | |
| " 'enthusiasm',\n", | |
| " 'peers',\n", | |
| " 'high school',\n", | |
| " 'fresh new crisp learning',\n", | |
| " 'lackluster movie theatre',\n", | |
| " 'renowned professor',\n", | |
| " 'another good book',\n", | |
| " 'friends',\n", | |
| " 'new opportunities',\n", | |
| " 'days',\n", | |
| " 'day',\n", | |
| " 'week',\n", | |
| " 'time',\n", | |
| " 'days week',\n", | |
| " 'hunger',\n", | |
| " 'weekend',\n", | |
| " 'textbook',\n", | |
| " 'university',\n", | |
| " 'another trip',\n", | |
| " 'breakfast',\n", | |
| " 'six classes',\n", | |
| " 'three meals',\n", | |
| " 'eager anticipation',\n", | |
| " 'us history',\n", | |
| " 'many factors',\n", | |
| " 'day',\n", | |
| " 'math science girl',\n", | |
| " 'formulas',\n", | |
| " 'fascination',\n", | |
| " 'later day',\n", | |
| " 'success',\n", | |
| " 'theorems',\n", | |
| " 'math justice',\n", | |
| " 'ancient greece',\n", | |
| " 'certain surface area',\n", | |
| " 'grocery store',\n", | |
| " 'real world',\n", | |
| " 'little change',\n", | |
| " 'pythagorean theorem',\n", | |
| " 'store',\n", | |
| " 'pascal triangle',\n", | |
| " 'golden rectangle',\n", | |
| " 'box',\n", | |
| " 'box',\n", | |
| " 'cumulative information',\n", | |
| " 'math',\n", | |
| " 'cashier',\n", | |
| " 'host',\n", | |
| " 'volume',\n", | |
| " 'problems',\n", | |
| " 'one day',\n", | |
| " 'fact',\n", | |
| " 'person',\n", | |
| " 'nearly 12 years',\n", | |
| " 'foreign symbols',\n", | |
| " 'simplicity',\n", | |
| " 'application',\n", | |
| " 'calculations',\n", | |
| " 'answer',\n", | |
| " 'genius',\n", | |
| " 'reasons',\n", | |
| " 'extra instruction',\n", | |
| " 'middle school',\n", | |
| " 'calculus ab',\n", | |
| " 'high low speed math class',\n", | |
| " 'slow speed math',\n", | |
| " 'review books',\n", | |
| " 'incoming 6th graders',\n", | |
| " 'junior year',\n", | |
| " '5th grade',\n", | |
| " 'classes',\n", | |
| " 'grade',\n", | |
| " 'second day',\n", | |
| " 'nearly year calculus',\n", | |
| " 'result',\n", | |
| " 'older brother',\n", | |
| " 'challenge',\n", | |
| " 'c',\n", | |
| " 'b',\n", | |
| " 'nearly year',\n", | |
| " 'c',\n", | |
| " 'scores',\n", | |
| " 'lot',\n", | |
| " 'parents',\n", | |
| " 'time',\n", | |
| " 'course',\n", | |
| " 'persistence',\n", | |
| " 'strongest subject',\n", | |
| " 'test',\n", | |
| " '',\n", | |
| " 'b',\n", | |
| " 'dedication',\n", | |
| " 'summer',\n", | |
| " 'transient love',\n", | |
| " 'love',\n", | |
| " 'love',\n", | |
| " 'good times',\n", | |
| " 'four years old',\n", | |
| " 'language',\n", | |
| " 'side',\n", | |
| " 'person',\n", | |
| " 'language',\n", | |
| " 'western influence',\n", | |
| " 'innocent ear',\n", | |
| " 'sweet tender words',\n", | |
| " 'thing',\n", | |
| " 'iron curtain',\n", | |
| " 'perfectly articulated phrases',\n", | |
| " 'opinion',\n", | |
| " 'country',\n", | |
| " 'matter',\n", | |
| " 'truth',\n", | |
| " 'foreign',\n", | |
| " 'especially family',\n", | |
| " 'spanish language telenovelas',\n", | |
| " 'romanian audiences',\n", | |
| " 'soap operas',\n", | |
| " 'new aspects',\n", | |
| " 'romanian cable network',\n", | |
| " 'language',\n", | |
| " 'june 16th',\n", | |
| " 'local cable company',\n", | |
| " 'language',\n", | |
| " 'romanian subtitles',\n", | |
| " 'spanish dialogue',\n", | |
| " 'united states',\n", | |
| " 'living room',\n", | |
| " 'little girl',\n", | |
| " 'world',\n", | |
| " 'rest',\n", | |
| " 'one',\n", | |
| " 'nonsense',\n", | |
| " 'june 16th 1994',\n", | |
| " 'man',\n", | |
| " 'apparent obsession',\n", | |
| " 'quite accomplishment',\n", | |
| " 'door',\n", | |
| " 'time',\n", | |
| " 'father',\n", | |
| " 'little interaction',\n", | |
| " 'los angeles',\n", | |
| " 'many hardships',\n", | |
| " 'huge indispensable part',\n", | |
| " 'language',\n", | |
| " 'spanish speaker paradise',\n", | |
| " 'largely romanian neighborhood',\n", | |
| " 'feelings',\n", | |
| " 'us',\n", | |
| " 'family objection',\n", | |
| " 'immigration',\n", | |
| " 'six years',\n", | |
| " 'advanced placement spanish',\n", | |
| " 'advanced placement',\n", | |
| " 'high school',\n", | |
| " 'new era',\n", | |
| " 'language',\n", | |
| " 'era',\n", | |
| " 'life',\n", | |
| " 'love',\n", | |
| " 'hour',\n", | |
| " 'two years',\n", | |
| " 'great asset',\n", | |
| " 'greatest compliment',\n", | |
| " 'second spoken language',\n", | |
| " 'language',\n", | |
| " 'romanian last name',\n", | |
| " 'objections',\n", | |
| " 'whose family members',\n", | |
| " 'natives',\n", | |
| " 'romanian girl',\n", | |
| " 'ability',\n", | |
| " 'husband',\n", | |
| " 'recent trip',\n", | |
| " 'united states',\n", | |
| " 'father',\n", | |
| " 'one man',\n", | |
| " 'words',\n", | |
| " 'unstoppable perseverance',\n", | |
| " 'family objections',\n", | |
| " 'whose native language',\n", | |
| " 'primary language',\n", | |
| " 'professional personal level',\n", | |
| " 'ambitious aggressive spirit',\n", | |
| " 'united states',\n", | |
| " '21 nations',\n", | |
| " 'nations',\n", | |
| " 'six official languages',\n", | |
| " 'new land',\n", | |
| " 'face',\n", | |
| " 'un',\n", | |
| " 'world',\n", | |
| " 'globe',\n", | |
| " 'part',\n", | |
| " 'people',\n", | |
| " 'link',\n", | |
| " 'impact',\n", | |
| " 'love',\n", | |
| " 'love',\n", | |
| " 'ability',\n", | |
| " '40 million americans',\n", | |
| " 'fight',\n", | |
| " 'path',\n", | |
| " '40 million',\n", | |
| " 'passion',\n", | |
| " 'los angeles',\n", | |
| " 'native adoptive countries',\n", | |
| " 'big part',\n", | |
| " 'language',\n", | |
| " 'even bigger part',\n", | |
| " 'remarkable accomplishments',\n", | |
| " 'romanian american girl',\n", | |
| " 'future',\n", | |
| " 'rest',\n", | |
| " 'case',\n", | |
| " 'years',\n", | |
| " 'teenage female angst',\n", | |
| " 'valuable information',\n", | |
| " 'holden caulfield',\n", | |
| " 'everyday occurrences',\n", | |
| " 'seemingly mindless trashy zombie films',\n", | |
| " 'clean thoughts',\n", | |
| " 'sexual imagery',\n", | |
| " 'profane vulgar language',\n", | |
| " 'honors program',\n", | |
| " 'dirty words',\n", | |
| " 'vampire slayer',\n", | |
| " 'subtle transcendent moments',\n", | |
| " 'madame bovary',\n", | |
| " 'cracked kettle',\n", | |
| " 'course',\n", | |
| " 'best brightest students',\n", | |
| " 'six week intensive college like experience',\n", | |
| " 'best thing',\n", | |
| " 'stars',\n", | |
| " 'exercise',\n", | |
| " 'governor honors program',\n", | |
| " 'brand new way',\n", | |
| " 'classes',\n", | |
| " 'class',\n", | |
| " 'classes',\n", | |
| " 'experiences',\n", | |
| " 'classics',\n", | |
| " 'bears',\n", | |
| " 'use',\n", | |
| " 'varying education',\n", | |
| " 'world',\n", | |
| " 'hackneyed phrase',\n", | |
| " 'extremes',\n", | |
| " 'seventeen year olds',\n", | |
| " 'close confines',\n", | |
| " 'mind',\n", | |
| " 'sixteen seventeen year olds',\n", | |
| " 'summer',\n", | |
| " 'teachers',\n", | |
| " 'thing',\n", | |
| " 'things',\n", | |
| " 'classroom',\n", | |
| " 'classrooms',\n", | |
| " 'people',\n", | |
| " 'interesting perspective',\n", | |
| " 'world',\n", | |
| " 'big name',\n", | |
| " 'best friends',\n", | |
| " 'running joke',\n", | |
| " 'growth',\n", | |
| " 'ethical framework',\n", | |
| " 'accomplished singer',\n", | |
| " 'endearing trick',\n", | |
| " 'beliefs',\n", | |
| " 'former girl friend',\n", | |
| " 'excellent athlete',\n", | |
| " 'phenomenal writer',\n", | |
| " 'clashes',\n", | |
| " 'math major',\n", | |
| " 'fairly conservative christians',\n", | |
| " 'every sport',\n", | |
| " 'sort',\n", | |
| " 'lot',\n", | |
| " 'experience',\n", | |
| " 'college undergrads',\n", | |
| " 'minority introduction',\n", | |
| " '68 rising seniors',\n", | |
| " 'highly rigorous academic program',\n", | |
| " 'small auditorium',\n", | |
| " 'program',\n", | |
| " 'lives',\n", | |
| " 'past summer',\n", | |
| " 'classes',\n", | |
| " 'top',\n", | |
| " 'opportunity',\n", | |
| " 'minds',\n", | |
| " 'confidence',\n", | |
| " 'six half weeks',\n", | |
| " 'hard times',\n", | |
| " 'problem sets',\n", | |
| " 'drill motors',\n", | |
| " 'best friends',\n", | |
| " 'time',\n", | |
| " 'mites valedictorians',\n", | |
| " 'support system',\n", | |
| " 'semiautonomous robot',\n", | |
| " 'even family',\n", | |
| " 'strangers',\n", | |
| " 'second week',\n", | |
| " 'shoulders',\n", | |
| " 'morning',\n", | |
| " 'book per week',\n", | |
| " 'book',\n", | |
| " '16 hours days',\n", | |
| " 'four hours',\n", | |
| " 'two morning',\n", | |
| " 'well known fact',\n", | |
| " 'average',\n", | |
| " 'ap physics',\n", | |
| " 'introductory level engineering',\n", | |
| " 'first year',\n", | |
| " 'ap literature',\n", | |
| " 'second year',\n", | |
| " 'introductory physics',\n", | |
| " 'engineering design',\n", | |
| " 'introductory physics engineering design',\n", | |
| " 'physics chemistry',\n", | |
| " 'full year',\n", | |
| " 'physics experience',\n", | |
| " '',\n", | |
| " 'semester',\n", | |
| " 'material',\n", | |
| " 'fall',\n", | |
| " 'schedule',\n", | |
| " '5 classes',\n", | |
| " 'half',\n", | |
| " 'work',\n", | |
| " 'month',\n", | |
| " 'month half',\n", | |
| " 'folklorico dancing',\n", | |
| " 'science bowl team',\n", | |
| " 'swim team',\n", | |
| " 'ace mentoring team',\n", | |
| " 'teams',\n", | |
| " 'educational opportunities',\n", | |
| " 'jewish student union',\n", | |
| " 'california honors society',\n", | |
| " 'play production',\n", | |
| " 'scholarship federation',\n", | |
| " 'science bowl',\n", | |
| " 'gear mentoring program',\n", | |
| " 'california honors society scholarship federation',\n", | |
| " 'new programs',\n", | |
| " 'academic school year',\n", | |
| " 'speech debate team',\n", | |
| " 'school',\n", | |
| " 'regular coursework',\n", | |
| " 'peak',\n", | |
| " 'involvement',\n", | |
| " 'program',\n", | |
| " 'lookout',\n", | |
| " 'academic endeavors',\n", | |
| " 'academic pursuits',\n", | |
| " 'bar',\n", | |
| " 'life',\n", | |
| " 'single pivotal point',\n", | |
| " 'challenging experience',\n", | |
| " 'program',\n", | |
| " 'program',\n", | |
| " 'assistants',\n", | |
| " 'motivation',\n", | |
| " 'breath',\n", | |
| " 'eyes',\n", | |
| " 'apology',\n", | |
| " 'incoherent words',\n", | |
| " 'adjacent words',\n", | |
| " 'simple word',\n", | |
| " 'familiar word',\n", | |
| " 'last word',\n", | |
| " 'sixth word',\n", | |
| " 'word',\n", | |
| " 'one word',\n", | |
| " 'incomplete translation',\n", | |
| " 'coherent translation',\n", | |
| " 'important word',\n", | |
| " 'deep breath',\n", | |
| " 'apology',\n", | |
| " 'clause',\n", | |
| " 'senses',\n", | |
| " 'mercy',\n", | |
| " '20 lines',\n", | |
| " 'initial scan',\n", | |
| " 'greek letters',\n", | |
| " 'bell rings',\n", | |
| " 'text',\n", | |
| " 'page',\n", | |
| " 'head',\n", | |
| " 'lexicon',\n", | |
| " 'eyes',\n", | |
| " 'whirlwind',\n", | |
| " 'figure skating',\n", | |
| " 'different mediums',\n", | |
| " 'hard work',\n", | |
| " 'value',\n", | |
| " 'apology',\n", | |
| " 'cold wind',\n", | |
| " 'thin sheets',\n", | |
| " 'new bruise',\n", | |
| " 'body',\n", | |
| " 'body',\n", | |
| " 'music',\n", | |
| " 'next jump',\n", | |
| " 'eyes',\n", | |
| " 'sharp skid',\n", | |
| " 'next program',\n", | |
| " 'air',\n", | |
| " 'visual performance',\n", | |
| " 'three times',\n", | |
| " 'bright january morning',\n", | |
| " 'jump',\n", | |
| " 'ice',\n", | |
| " 'knees',\n", | |
| " 'face',\n", | |
| " 'translation',\n", | |
| " 'torso',\n", | |
| " 'arms',\n", | |
| " 'every note',\n", | |
| " 'every pitch',\n", | |
| " 'cadenza',\n", | |
| " 'ground',\n", | |
| " 'shock',\n", | |
| " 'midst',\n", | |
| " 'frame',\n", | |
| " 'meadow lake',\n", | |
| " 'proper removal process',\n", | |
| " 'mugwort removal cleanup',\n", | |
| " 'strenuous month',\n", | |
| " 'tedious logistical process',\n", | |
| " 'discovery guide leader',\n", | |
| " 'july 2018',\n", | |
| " 'permit',\n", | |
| " 'importance',\n", | |
| " 'activism',\n", | |
| " 'plan',\n", | |
| " 'proper removal technique',\n", | |
| " 'eager group',\n", | |
| " 'invasive species',\n", | |
| " 'productive plan',\n", | |
| " 'shore',\n", | |
| " 'knowledge',\n", | |
| " 'soil',\n", | |
| " 'stem',\n", | |
| " 'person',\n", | |
| " 'shovel',\n", | |
| " 'hot sun',\n", | |
| " 'back',\n", | |
| " 'plant',\n", | |
| " '4 2 pounds',\n", | |
| " 'benefits',\n", | |
| " 'hold',\n", | |
| " 'ground',\n", | |
| " 'twenty pairs',\n", | |
| " 'day',\n", | |
| " 'plans',\n", | |
| " 'songs',\n", | |
| " 'many things',\n", | |
| " 'actions',\n", | |
| " 'plato years ago',\n", | |
| " 'knowledge',\n", | |
| " 'lessons',\n", | |
| " 'world',\n", | |
| " 'apology',\n", | |
| " 'math packet',\n", | |
| " 'mental contract',\n", | |
| " 'pages',\n", | |
| " 'packet',\n", | |
| " 'black italicized letters',\n", | |
| " 'life',\n", | |
| " 'arms',\n", | |
| " 'top bunk',\n", | |
| " 'challenge',\n", | |
| " 'pencil',\n", | |
| " 'hand',\n", | |
| " 'title',\n", | |
| " 'problems',\n", | |
| " 'goosebumps',\n", | |
| " 'mysteries',\n", | |
| " 'words',\n", | |
| " 'teacher',\n", | |
| " 'fingers',\n", | |
| " 'cover',\n", | |
| " 'better pay',\n", | |
| " 'minimum wage job',\n", | |
| " 'makeshift bed',\n", | |
| " 'second bed',\n", | |
| " 'torn scratchy old sofa',\n", | |
| " 'bed',\n", | |
| " 'one bed',\n", | |
| " 'creative one',\n", | |
| " 'new home',\n", | |
| " 'one job',\n", | |
| " 'little room',\n", | |
| " 'experiences',\n", | |
| " 'omnipresent challenge',\n", | |
| " 'many challenges',\n", | |
| " 'dad',\n", | |
| " 'dad',\n", | |
| " 'mom',\n", | |
| " 'mom',\n", | |
| " 'pursuit',\n", | |
| " 'another challenge',\n", | |
| " 'hard work',\n", | |
| " 'bunk',\n", | |
| " 'family',\n", | |
| " 'atlantic',\n", | |
| " 'life',\n", | |
| " 'late working',\n", | |
| " 'first year',\n", | |
| " 'crib',\n", | |
| " 'belongings',\n", | |
| " 'comfort',\n", | |
| " 'tenacity',\n", | |
| " 'consequences',\n", | |
| " 'love',\n", | |
| " 'parents',\n", | |
| " 'sister',\n", | |
| " 'day',\n", | |
| " 'one example',\n", | |
| " 'ten years ago',\n", | |
| " 'another stage',\n", | |
| " 'value',\n", | |
| " 'values',\n", | |
| " 'single job',\n", | |
| " 'dad consistent job switches',\n", | |
| " 'mom devotion',\n", | |
| " 'family',\n", | |
| " 'sisters',\n", | |
| " 'challenge',\n", | |
| " 'packet',\n", | |
| " 'grueling two months',\n", | |
| " 'two months',\n", | |
| " 'experiences',\n", | |
| " 'questions',\n", | |
| " 'parents',\n", | |
| " '11 year old self',\n", | |
| " 'ten years',\n", | |
| " 'middle high school',\n", | |
| " 'learned values',\n", | |
| " 'school',\n", | |
| " 'goals',\n", | |
| " 'goal',\n", | |
| " 'value',\n", | |
| " 'values',\n", | |
| " 'challenging courses',\n", | |
| " 'past five years',\n", | |
| " 'guidance',\n", | |
| " 'people',\n", | |
| " 'surgeon',\n", | |
| " 'career',\n", | |
| " 'church',\n", | |
| " 'person',\n", | |
| " 'way',\n", | |
| " 'community',\n", | |
| " 'time',\n", | |
| " 'muddy soccer cleats',\n", | |
| " 'high grades',\n", | |
| " 'successes',\n", | |
| " 'results',\n", | |
| " 'field hockey gear',\n", | |
| " '15 field hockey team',\n", | |
| " 'wooden bench',\n", | |
| " 'cleats',\n", | |
| " 'locker',\n", | |
| " 'heavy legs',\n", | |
| " 'wet dreary october evening',\n", | |
| " 'concrete',\n", | |
| " 'netherlands',\n", | |
| " 'homework',\n", | |
| " 'year',\n", | |
| " 'practice',\n", | |
| " 'head',\n", | |
| " 'dirt',\n", | |
| " 'sound',\n", | |
| " 'field hockey',\n", | |
| " 'field hockey cleats',\n", | |
| " 'new opportunities',\n", | |
| " 'deerfield field hockey coach',\n", | |
| " 'gender boundaries',\n", | |
| " 'girls field hockey practice',\n", | |
| " 'tough transition',\n", | |
| " 'eyes',\n", | |
| " 'warm rush',\n", | |
| " 'ms mcvaugh',\n", | |
| " 'unusual opportunity',\n", | |
| " 'red faces',\n", | |
| " 'gender barrier',\n", | |
| " 'shared fanaticism',\n", | |
| " 'bench',\n", | |
| " 'clamor',\n", | |
| " 'burning glow',\n", | |
| " 'initial anxiety',\n", | |
| " 'exuberance',\n", | |
| " 'girls',\n", | |
| " 'veins',\n", | |
| " 'following day',\n", | |
| " 'breath',\n", | |
| " 'emotion',\n", | |
| " 'second',\n", | |
| " 'body',\n", | |
| " 'thought',\n", | |
| " 'warm',\n", | |
| " 'step',\n", | |
| " 'ball',\n", | |
| " 'turf',\n", | |
| " 'answer',\n", | |
| " 'moment',\n", | |
| " 'end',\n", | |
| " 'whistle',\n", | |
| " 'new skills',\n", | |
| " 'new athletic challenges',\n", | |
| " 'new approach',\n", | |
| " 'soccer field',\n", | |
| " 'new improvement',\n", | |
| " 'sport',\n", | |
| " 'atmosphere',\n", | |
| " 'value',\n", | |
| " 'experience',\n", | |
| " 'joy',\n", | |
| " 'turf',\n", | |
| " 'focus',\n", | |
| " 'successes',\n", | |
| " 'moments',\n", | |
| " 'new material',\n", | |
| " 'boarding school',\n", | |
| " 'field hockey gear',\n", | |
| " 'organic chemistry',\n", | |
| " 'new mindset',\n", | |
| " 'deeper questions',\n", | |
| " 'even athletic achievements',\n", | |
| " 'opportunities',\n", | |
| " 'dutch stroopwafels',\n", | |
| " 'open curious risk taking mindset',\n", | |
| " 'nucleophile electrophile reactions',\n", | |
| " 'exploration',\n", | |
| " 'grades',\n", | |
| " 'cultural differences',\n", | |
| " 'hall',\n", | |
| " 'peers',\n", | |
| " 'result',\n", | |
| " 'classroom',\n", | |
| " 'dorm',\n", | |
| " 'opportunities',\n", | |
| " 'experiences',\n", | |
| " 'obstacles',\n", | |
| " 'successes',\n", | |
| " 'dorm',\n", | |
| " 'classroom',\n", | |
| " 'cleats',\n", | |
| " 'heart',\n", | |
| " 'lingering poor taste',\n", | |
| " 'creamy white center',\n", | |
| " 'mouth',\n", | |
| " 'firm chocolate crust',\n", | |
| " 'oreo',\n", | |
| " 'social meaning',\n", | |
| " 'exterior',\n", | |
| " 'confused stares',\n", | |
| " 'new story',\n", | |
| " 'hans zimmer',\n", | |
| " 'hans zimmer',\n", | |
| " 'high school',\n", | |
| " 'top 40 hits',\n", | |
| " 'note',\n", | |
| " 'little girl',\n", | |
| " 'black peers',\n", | |
| " 'upbeat peppy nature',\n", | |
| " 'tango melody',\n", | |
| " 'musical taste',\n", | |
| " 'vastly different styles',\n", | |
| " 'ballroom pop music',\n", | |
| " 'seductive powerful attacks',\n", | |
| " 'whitest thing',\n", | |
| " 'friend',\n", | |
| " 'oreo',\n", | |
| " 'new guilty sensations',\n", | |
| " 'white people',\n", | |
| " 'social alienation',\n", | |
| " 'black roots',\n", | |
| " 'black person',\n", | |
| " 'skin tone',\n", | |
| " 'african american woman',\n", | |
| " 'certain race',\n", | |
| " 'ballroom pop music',\n", | |
| " 'negatively charged implications',\n", | |
| " 'called white core',\n", | |
| " 'musical interests',\n", | |
| " 'expectations',\n", | |
| " 'oreo',\n", | |
| " 'two seemingly different commodities',\n", | |
| " 'connection',\n", | |
| " 'various musical styles',\n", | |
| " 'county solo ensemble festival',\n", | |
| " 'lyrical composition',\n", | |
| " 'day',\n", | |
| " 'band director',\n", | |
| " 'page',\n", | |
| " 'solo',\n", | |
| " 'clarinet player',\n", | |
| " 'rhythms',\n", | |
| " 'band',\n", | |
| " 'freshman year',\n", | |
| " 'life',\n", | |
| " 'notes',\n", | |
| " 'piece',\n", | |
| " 'right time',\n", | |
| " 'right note',\n", | |
| " 'piece',\n", | |
| " 'interpretation',\n", | |
| " 'musical markings',\n", | |
| " 'breath marks',\n", | |
| " 'black ink',\n", | |
| " 'white space',\n", | |
| " 'climactic precipice',\n", | |
| " 'contribution',\n", | |
| " 'first glance',\n", | |
| " 'second glance',\n", | |
| " 'black ink',\n", | |
| " 'dynamic journey',\n", | |
| " 'definitive nature',\n", | |
| " 'measures',\n", | |
| " 'pencil',\n", | |
| " 'sheet',\n", | |
| " 'composition',\n", | |
| " 'composition',\n", | |
| " 'audience',\n", | |
| " 'distinction',\n", | |
| " 'rhythms',\n", | |
| " 'realization',\n", | |
| " 'notes',\n", | |
| " 'certain races',\n", | |
| " 'stereotypes',\n", | |
| " 'documented definition',\n", | |
| " 'mind',\n", | |
| " 'minds',\n", | |
| " 'oreo',\n", | |
| " 'black woman society',\n", | |
| " 'proud black woman',\n", | |
| " 'ballroom music',\n", | |
| " 'black experience',\n", | |
| " 'music maker',\n", | |
| " 'input',\n", | |
| " 'beautiful addition',\n", | |
| " 'design',\n", | |
| " 'blank canvas',\n", | |
| " 'creamy white center',\n", | |
| " 'ever evolving character',\n", | |
| " 'mark',\n", | |
| " 'one direction tune',\n", | |
| " 'betrayal',\n", | |
| " 'musical interests',\n", | |
| " 'definition',\n", | |
| " 'desire',\n", | |
| " 'roots',\n", | |
| " 'grip',\n", | |
| " 'intricacies',\n", | |
| " 'oreo',\n", | |
| " 'choice',\n", | |
| " 'style',\n", | |
| " 'parents',\n", | |
| " 'clock',\n", | |
| " 'letter',\n", | |
| " 'sibling',\n", | |
| " 'friends',\n", | |
| " 'lhasa apso dog',\n", | |
| " 'child',\n", | |
| " 'new sister',\n", | |
| " 'whole',\n", | |
| " 'ways',\n", | |
| " 'dog',\n", | |
| " 'life',\n", | |
| " 'community',\n", | |
| " 'group',\n", | |
| " 'parents',\n", | |
| " 'furry',\n", | |
| " 'sentiments',\n", | |
| " 'world history classes',\n", | |
| " 'east asian',\n", | |
| " 'vast differences',\n", | |
| " 'less importance',\n", | |
| " 'social studies classes',\n", | |
| " 'european east asian history',\n", | |
| " 'cheese sandwiches',\n", | |
| " 'filipino girl',\n", | |
| " 'way',\n", | |
| " 'traditional adobo chicken',\n", | |
| " 'immigrant parents',\n", | |
| " 'pathway',\n", | |
| " 'filipino self',\n", | |
| " 'filipino delicacies',\n", | |
| " 'huge impact',\n", | |
| " 'cultures',\n", | |
| " 'friends',\n", | |
| " '75 white students',\n", | |
| " 'community',\n", | |
| " 'way',\n", | |
| " 'norm',\n", | |
| " 'grade',\n", | |
| " 'mom',\n", | |
| " 'ancestry',\n", | |
| " 'one category',\n", | |
| " 'predominantly white town',\n", | |
| " 'pungent aromas',\n", | |
| " 'population',\n", | |
| " 'school',\n", | |
| " 'heritage',\n", | |
| " '75',\n", | |
| " 'notion',\n", | |
| " 'classmates',\n", | |
| " 'different viewpoints',\n", | |
| " 'young conservatives club',\n", | |
| " 'current state',\n", | |
| " 'high school',\n", | |
| " 'high school democrats club young conservatives club',\n", | |
| " 'high school democrats club',\n", | |
| " 'filipino dishes',\n", | |
| " 'filipino pride',\n", | |
| " 'social studies',\n", | |
| " 'freshman year',\n", | |
| " 'student body',\n", | |
| " 'feelings',\n", | |
| " 'school',\n", | |
| " 'philippines',\n", | |
| " 'speech debate tournaments',\n", | |
| " 'background',\n", | |
| " 'history',\n", | |
| " 'increase',\n", | |
| " 'spectrum',\n", | |
| " 'globe',\n", | |
| " 'array',\n", | |
| " 'friends',\n", | |
| " 'peers',\n", | |
| " 'message',\n", | |
| " 'filipino background',\n", | |
| " 'appreciation',\n", | |
| " 'norm',\n", | |
| " 'potential',\n", | |
| " 'community',\n", | |
| " 'narrative',\n", | |
| " 'conventional classrooms',\n", | |
| " 'following day',\n", | |
| " 'past summer',\n", | |
| " 'test',\n", | |
| " 'things',\n", | |
| " 'paper planes',\n", | |
| " 'new jersey',\n", | |
| " 'water',\n", | |
| " 'solar ovens',\n", | |
| " 'light space vacuums',\n", | |
| " 'best brightest students',\n", | |
| " 'physical science',\n", | |
| " '200 students',\n", | |
| " 'new jersey',\n", | |
| " 'topic',\n", | |
| " 'speed',\n", | |
| " 'content',\n", | |
| " 'next weeks',\n", | |
| " 'dynamic',\n", | |
| " 'college campus',\n", | |
| " 'spk program',\n", | |
| " 'box',\n", | |
| " 'box',\n", | |
| " 'five week enrichment program',\n", | |
| " 'top',\n", | |
| " 'textbooks',\n", | |
| " 'table',\n", | |
| " 'first day',\n", | |
| " 'siphon',\n", | |
| " 'phones',\n", | |
| " 'teacher',\n", | |
| " 'brains',\n", | |
| " 'impressive terminology',\n", | |
| " 'ap physics exam',\n", | |
| " 'ap physics',\n", | |
| " 'better future',\n", | |
| " 'community',\n", | |
| " 'competition',\n", | |
| " 'problems',\n", | |
| " 'means',\n", | |
| " 'five weeks',\n", | |
| " 'many tropical diseases',\n", | |
| " 'yellow fever',\n", | |
| " 'yellow fever',\n", | |
| " 'blood tests',\n", | |
| " 'several hours',\n", | |
| " 'urban hospital',\n", | |
| " 'skin',\n", | |
| " 'local doctor',\n", | |
| " 'time',\n", | |
| " 'diagnosis',\n", | |
| " 'solvable issue',\n", | |
| " 'family',\n", | |
| " 'small changes',\n", | |
| " 'neighboring villages',\n", | |
| " 'health education programs',\n", | |
| " 'organizations',\n", | |
| " 'immense impacts',\n", | |
| " 'several universities',\n", | |
| " 'small stands',\n", | |
| " 'simple things',\n", | |
| " 'real life problems',\n", | |
| " 'medical clinics',\n", | |
| " 'water sanitation plants',\n", | |
| " 'third parties',\n", | |
| " 'branch president',\n", | |
| " 'public gatherings',\n", | |
| " 'non profit organization',\n", | |
| " 'health social issues',\n", | |
| " 'small village',\n", | |
| " 'lifestyle changes',\n", | |
| " 'big effect',\n", | |
| " 'basic needs',\n", | |
| " 'first high school branch',\n", | |
| " 'one village',\n", | |
| " ...]" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 20 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "yDfFHlLmtxVd" | |
| }, | |
| "source": [ | |
| "import tensorflow as tf\n", | |
| "import pickle \n", | |
| "import numpy as np \n", | |
| "from tensorflow.keras.layers import Embedding, LSTM, Dense\n", | |
| "from tensorflow.keras.models import Sequential\n", | |
| "from tensorflow.keras.utils import to_categorical\n", | |
| "from tensorflow.keras.optimizers import Adam\n", | |
| "\n", | |
| "tokenizer = tf.keras.preprocessing.text.Tokenizer()\n", | |
| "tokenizer.fit_on_texts([text]) # building the tokenizer\n", | |
| "\n", | |
| "pickle.dump(tokenizer, open('tokenizer1.pkl', 'wb')) # can be downloaded to run predictions without training\n", | |
| "\n", | |
| "sequence_data = tokenizer.texts_to_sequences([text])[0] # building sequences\n", | |
| "\n", | |
| "vocab_size = len(tokenizer.word_index) + 1\n", | |
| "\n", | |
| "sequences = [] # for words from sequence data\n", | |
| "\n", | |
| "for i in range(1, len(sequence_data)):\n", | |
| " words = sequence_data[i-1:i+1]\n", | |
| " sequences.append(words)\n", | |
| " \n", | |
| "sequences = np.array(sequences)\n", | |
| "\n", | |
| "X = []\n", | |
| "y = []\n", | |
| "\n", | |
| "for i in sequences:\n", | |
| " X.append(i[0])\n", | |
| " y.append(i[1])\n", | |
| " \n", | |
| "X = np.array(X)\n", | |
| "y = np.array(y)\n", | |
| "\n", | |
| "y = to_categorical(y, num_classes=vocab_size)" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "XdX5jXp4wXrN", | |
| "outputId": "9183ecc8-4aba-4ab5-d009-ff49f27ed61b" | |
| }, | |
| "source": [ | |
| "model = tf.keras.models.Sequential()\n", | |
| "model.add(Embedding(vocab_size, 10, input_length=1))\n", | |
| "model.add(LSTM(1000, return_sequences=True))\n", | |
| "model.add(LSTM(1000))\n", | |
| "model.add(Dense(1000, activation=\"relu\"))\n", | |
| "model.add(Dense(vocab_size, activation=\"softmax\"))\n", | |
| "\n", | |
| "model.summary()" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Model: \"sequential\"\n", | |
| "_________________________________________________________________\n", | |
| "Layer (type) Output Shape Param # \n", | |
| "=================================================================\n", | |
| "embedding (Embedding) (None, 1, 10) 63720 \n", | |
| "_________________________________________________________________\n", | |
| "lstm (LSTM) (None, 1, 1000) 4044000 \n", | |
| "_________________________________________________________________\n", | |
| "lstm_1 (LSTM) (None, 1000) 8004000 \n", | |
| "_________________________________________________________________\n", | |
| "dense (Dense) (None, 1000) 1001000 \n", | |
| "_________________________________________________________________\n", | |
| "dense_1 (Dense) (None, 6372) 6378372 \n", | |
| "=================================================================\n", | |
| "Total params: 19,491,092\n", | |
| "Trainable params: 19,491,092\n", | |
| "Non-trainable params: 0\n", | |
| "_________________________________________________________________\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "E9UMpge3w-SU" | |
| }, | |
| "source": [ | |
| "from tensorflow.keras.callbacks import ModelCheckpoint\n", | |
| "from tensorflow.keras.callbacks import ReduceLROnPlateau\n", | |
| "from tensorflow.keras.callbacks import TensorBoard\n", | |
| "\n", | |
| "checkpoint = ModelCheckpoint(\"nextword1.h5\", monitor='loss', verbose=1,\n", | |
| " save_best_only=True, mode='auto') # Checking loss after each epoch\n", | |
| "\n", | |
| "reduce = ReduceLROnPlateau(monitor='loss', factor=0.2, patience=3, min_lr=0.0001, verbose = 1) # For optimization of loss\n" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "UWKkN8vtxCLA" | |
| }, | |
| "source": [ | |
| "model.compile(loss=\"categorical_crossentropy\", optimizer=Adam(learning_rate=0.001))" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "I-Zv6sfGxFEm", | |
| "outputId": "21cdccb6-4b89-4c29-d746-b66a759f3ac8" | |
| }, | |
| "source": [ | |
| "model.fit(X, y, epochs=50, batch_size=32, callbacks=[checkpoint, reduce])\n", | |
| "# Loss goes smoothly from 8.5 to 1.3" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Epoch 1/50\n", | |
| "310/310 [==============================] - 13s 14ms/step - loss: 8.7378\n", | |
| "\n", | |
| "Epoch 00001: loss improved from inf to 8.73776, saving model to nextword1.h5\n", | |
| "Epoch 2/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 8.4606\n", | |
| "\n", | |
| "Epoch 00002: loss improved from 8.73776 to 8.46062, saving model to nextword1.h5\n", | |
| "Epoch 3/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 8.1330\n", | |
| "\n", | |
| "Epoch 00003: loss improved from 8.46062 to 8.13299, saving model to nextword1.h5\n", | |
| "Epoch 4/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 7.6964\n", | |
| "\n", | |
| "Epoch 00004: loss improved from 8.13299 to 7.69636, saving model to nextword1.h5\n", | |
| "Epoch 5/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 7.1067\n", | |
| "\n", | |
| "Epoch 00005: loss improved from 7.69636 to 7.10675, saving model to nextword1.h5\n", | |
| "Epoch 6/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 6.4593\n", | |
| "\n", | |
| "Epoch 00006: loss improved from 7.10675 to 6.45932, saving model to nextword1.h5\n", | |
| "Epoch 7/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 5.8750\n", | |
| "\n", | |
| "Epoch 00007: loss improved from 6.45932 to 5.87500, saving model to nextword1.h5\n", | |
| "Epoch 8/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 5.4042\n", | |
| "\n", | |
| "Epoch 00008: loss improved from 5.87500 to 5.40424, saving model to nextword1.h5\n", | |
| "Epoch 9/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 4.9850\n", | |
| "\n", | |
| "Epoch 00009: loss improved from 5.40424 to 4.98504, saving model to nextword1.h5\n", | |
| "Epoch 10/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 4.5939\n", | |
| "\n", | |
| "Epoch 00010: loss improved from 4.98504 to 4.59394, saving model to nextword1.h5\n", | |
| "Epoch 11/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 4.2755\n", | |
| "\n", | |
| "Epoch 00011: loss improved from 4.59394 to 4.27551, saving model to nextword1.h5\n", | |
| "Epoch 12/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 3.9807\n", | |
| "\n", | |
| "Epoch 00012: loss improved from 4.27551 to 3.98068, saving model to nextword1.h5\n", | |
| "Epoch 13/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 3.7281\n", | |
| "\n", | |
| "Epoch 00013: loss improved from 3.98068 to 3.72815, saving model to nextword1.h5\n", | |
| "Epoch 14/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 3.4971\n", | |
| "\n", | |
| "Epoch 00014: loss improved from 3.72815 to 3.49705, saving model to nextword1.h5\n", | |
| "Epoch 15/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 3.2748\n", | |
| "\n", | |
| "Epoch 00015: loss improved from 3.49705 to 3.27483, saving model to nextword1.h5\n", | |
| "Epoch 16/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 3.0503\n", | |
| "\n", | |
| "Epoch 00016: loss improved from 3.27483 to 3.05034, saving model to nextword1.h5\n", | |
| "Epoch 17/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 2.8535\n", | |
| "\n", | |
| "Epoch 00017: loss improved from 3.05034 to 2.85349, saving model to nextword1.h5\n", | |
| "Epoch 18/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 2.6912\n", | |
| "\n", | |
| "Epoch 00018: loss improved from 2.85349 to 2.69123, saving model to nextword1.h5\n", | |
| "Epoch 19/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 2.5407\n", | |
| "\n", | |
| "Epoch 00019: loss improved from 2.69123 to 2.54072, saving model to nextword1.h5\n", | |
| "Epoch 20/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 2.4099\n", | |
| "\n", | |
| "Epoch 00020: loss improved from 2.54072 to 2.40990, saving model to nextword1.h5\n", | |
| "Epoch 21/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 2.2625\n", | |
| "\n", | |
| "Epoch 00021: loss improved from 2.40990 to 2.26249, saving model to nextword1.h5\n", | |
| "Epoch 22/50\n", | |
| "310/310 [==============================] - 4s 11ms/step - loss: 2.1525\n", | |
| "\n", | |
| "Epoch 00022: loss improved from 2.26249 to 2.15251, saving model to nextword1.h5\n", | |
| "Epoch 23/50\n", | |
| "310/310 [==============================] - 4s 11ms/step - loss: 2.0715\n", | |
| "\n", | |
| "Epoch 00023: loss improved from 2.15251 to 2.07147, saving model to nextword1.h5\n", | |
| "Epoch 24/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.9944\n", | |
| "\n", | |
| "Epoch 00024: loss improved from 2.07147 to 1.99440, saving model to nextword1.h5\n", | |
| "Epoch 25/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.9266\n", | |
| "\n", | |
| "Epoch 00025: loss improved from 1.99440 to 1.92657, saving model to nextword1.h5\n", | |
| "Epoch 26/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.8534\n", | |
| "\n", | |
| "Epoch 00026: loss improved from 1.92657 to 1.85343, saving model to nextword1.h5\n", | |
| "Epoch 27/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.8066\n", | |
| "\n", | |
| "Epoch 00027: loss improved from 1.85343 to 1.80657, saving model to nextword1.h5\n", | |
| "Epoch 28/50\n", | |
| "310/310 [==============================] - 4s 12ms/step - loss: 1.7287\n", | |
| "\n", | |
| "Epoch 00028: loss improved from 1.80657 to 1.72875, saving model to nextword1.h5\n", | |
| "Epoch 29/50\n", | |
| "310/310 [==============================] - 4s 12ms/step - loss: 1.7022\n", | |
| "\n", | |
| "Epoch 00029: loss improved from 1.72875 to 1.70218, saving model to nextword1.h5\n", | |
| "Epoch 30/50\n", | |
| "310/310 [==============================] - 4s 12ms/step - loss: 1.6743\n", | |
| "\n", | |
| "Epoch 00030: loss improved from 1.70218 to 1.67435, saving model to nextword1.h5\n", | |
| "Epoch 31/50\n", | |
| "310/310 [==============================] - 4s 11ms/step - loss: 1.6170\n", | |
| "\n", | |
| "Epoch 00031: loss improved from 1.67435 to 1.61702, saving model to nextword1.h5\n", | |
| "Epoch 32/50\n", | |
| "310/310 [==============================] - 4s 12ms/step - loss: 1.5930\n", | |
| "\n", | |
| "Epoch 00032: loss improved from 1.61702 to 1.59303, saving model to nextword1.h5\n", | |
| "Epoch 33/50\n", | |
| "310/310 [==============================] - 4s 12ms/step - loss: 1.5677\n", | |
| "\n", | |
| "Epoch 00033: loss improved from 1.59303 to 1.56768, saving model to nextword1.h5\n", | |
| "Epoch 34/50\n", | |
| "310/310 [==============================] - 4s 11ms/step - loss: 1.5256\n", | |
| "\n", | |
| "Epoch 00034: loss improved from 1.56768 to 1.52556, saving model to nextword1.h5\n", | |
| "Epoch 35/50\n", | |
| "310/310 [==============================] - 4s 11ms/step - loss: 1.5000\n", | |
| "\n", | |
| "Epoch 00035: loss improved from 1.52556 to 1.49998, saving model to nextword1.h5\n", | |
| "Epoch 36/50\n", | |
| "310/310 [==============================] - 4s 11ms/step - loss: 1.4897\n", | |
| "\n", | |
| "Epoch 00036: loss improved from 1.49998 to 1.48966, saving model to nextword1.h5\n", | |
| "Epoch 37/50\n", | |
| "310/310 [==============================] - 4s 11ms/step - loss: 1.4821\n", | |
| "\n", | |
| "Epoch 00037: loss improved from 1.48966 to 1.48205, saving model to nextword1.h5\n", | |
| "Epoch 38/50\n", | |
| "310/310 [==============================] - 4s 11ms/step - loss: 1.4629\n", | |
| "\n", | |
| "Epoch 00038: loss improved from 1.48205 to 1.46293, saving model to nextword1.h5\n", | |
| "Epoch 39/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.4266\n", | |
| "\n", | |
| "Epoch 00039: loss improved from 1.46293 to 1.42664, saving model to nextword1.h5\n", | |
| "Epoch 40/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.4134\n", | |
| "\n", | |
| "Epoch 00040: loss improved from 1.42664 to 1.41341, saving model to nextword1.h5\n", | |
| "Epoch 41/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.4186\n", | |
| "\n", | |
| "Epoch 00041: loss did not improve from 1.41341\n", | |
| "Epoch 42/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.4091\n", | |
| "\n", | |
| "Epoch 00042: loss improved from 1.41341 to 1.40906, saving model to nextword1.h5\n", | |
| "Epoch 43/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.3942\n", | |
| "\n", | |
| "Epoch 00043: loss improved from 1.40906 to 1.39421, saving model to nextword1.h5\n", | |
| "Epoch 44/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.3607\n", | |
| "\n", | |
| "Epoch 00044: loss improved from 1.39421 to 1.36070, saving model to nextword1.h5\n", | |
| "Epoch 45/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.3522\n", | |
| "\n", | |
| "Epoch 00045: loss improved from 1.36070 to 1.35223, saving model to nextword1.h5\n", | |
| "Epoch 46/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.3254\n", | |
| "\n", | |
| "Epoch 00046: loss improved from 1.35223 to 1.32537, saving model to nextword1.h5\n", | |
| "Epoch 47/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.3091\n", | |
| "\n", | |
| "Epoch 00047: loss improved from 1.32537 to 1.30907, saving model to nextword1.h5\n", | |
| "Epoch 48/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.2962\n", | |
| "\n", | |
| "Epoch 00048: loss improved from 1.30907 to 1.29615, saving model to nextword1.h5\n", | |
| "Epoch 49/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.3012\n", | |
| "\n", | |
| "Epoch 00049: loss did not improve from 1.29615\n", | |
| "Epoch 50/50\n", | |
| "310/310 [==============================] - 3s 11ms/step - loss: 1.3003\n", | |
| "\n", | |
| "Epoch 00050: loss did not improve from 1.29615\n" | |
| ], | |
| "name": "stdout" | |
| }, | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/plain": [ | |
| "<tensorflow.python.keras.callbacks.History at 0x7fc5fa5404d0>" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 25 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "DuItv7U2xQzN", | |
| "outputId": "f0a6160d-18eb-4819-d09e-600903691119" | |
| }, | |
| "source": [ | |
| "for i in range(3): # Running it 3 times\n", | |
| " text = input(\"Enter your line: \")\n", | |
| "\n", | |
| " text = text.split(\" \") # checking for multiple words\n", | |
| " text = text[-1] # taking last word\n", | |
| "\n", | |
| " text = ''.join(text)\n", | |
| " \n", | |
| " sequence = tokenizer.texts_to_sequences([text])[0] #building tokenised data for comparison\n", | |
| " sequence = np.array(sequence)\n", | |
| "\n", | |
| " preds = model.predict_classes(sequence)\n", | |
| "\n", | |
| " predicted_word = \"\"\n", | |
| " \n", | |
| " for key, value in tokenizer.word_index.items(): # finding the word corresponding to the predicted token\n", | |
| " if value == preds:\n", | |
| " predicted_word = key\n", | |
| " break\n", | |
| " \n", | |
| " print(f\"Predicted Next Word : {predicted_word}\")" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Enter your line: alarm clock\n" | |
| ], | |
| "name": "stdout" | |
| }, | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/sequential.py:455: UserWarning: `model.predict_classes()` is deprecated and will be removed after 2021-01-01. Please use instead:* `np.argmax(model.predict(x), axis=-1)`, if your model does multi-class classification (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`, if your model does binary classification (e.g. if it uses a `sigmoid` last-layer activation).\n", | |
| " warnings.warn('`model.predict_classes()` is deprecated and '\n" | |
| ], | |
| "name": "stderr" | |
| }, | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Predicted Next Word : steady beep\n", | |
| "Enter your line: high school\n", | |
| "Predicted Next Word : entrepreneur\n", | |
| "Enter your line: marvelous glorious thing\n", | |
| "Predicted Next Word : precious planet\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "lx4qh09_gN3B" | |
| }, | |
| "source": [ | |
| "**STEP 5**- Fine tuning word embedding (GloVe or Word2Vec)" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "cLyqvLyn-N3Z", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "outputId": "8a1d262f-e423-4feb-fcba-43de90f8ba99" | |
| }, | |
| "source": [ | |
| "sent = 'run python program run, to make this work' # Example of noun chunks\n", | |
| "pa = nlp(sent)\n", | |
| "l = list(pa.noun_chunks)\n", | |
| "[str(x) for x in l]" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/plain": [ | |
| "['python program run', 'this work']" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 85 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "93IZqt_vchz1" | |
| }, | |
| "source": [ | |
| "## Noun Chunks\n", | |
| "spacy returns Noun Chunks as type 'token.span.Span'\n", | |
| "\n", | |
| "We need it to be of type 'string' in order to build a word2vec \n", | |
| "\n", | |
| "**Do note that this vocabulary is quite small as compared to what is ideally required**" | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "ErpvoXs9gQiW", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "outputId": "0c84ccfd-a95e-4c31-9374-c2d40019adb8" | |
| }, | |
| "source": [ | |
| "#Fine tune word embedding model using the 'Noun' phrases in the data.\n", | |
| "\n", | |
| "#In this step, we will fine tune the word embedding model (either GloVe or Word2Vec) on the Nouns (use POS tag from spacy) from our hybrid database.\n", | |
| "\n", | |
| "#First, extract the nouns in the database in a sequential manner.\n", | |
| "nouns = []\n", | |
| "for i in sentences_list :\n", | |
| " parsed = nlp(i)\n", | |
| " temp = list(parsed.noun_chunks) \n", | |
| " nouns.append([str(x) for x in temp]) # take noun chunks and save them as string\n", | |
| "print(nouns[0])" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "['alarm', 'many high school students', 'monstrosity', 'whose purpose', 'torture sleep', 'situation', 'early risers', 'earsplitting noise', 'doubt']\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "UoQmPNm-8unK", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "outputId": "d4e68196-c768-4f99-c59f-831179a77ca0" | |
| }, | |
| "source": [ | |
| "!pip install gensim" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Requirement already satisfied: gensim in /usr/local/lib/python3.7/dist-packages (3.6.0)\n", | |
| "Requirement already satisfied: six>=1.5.0 in /usr/local/lib/python3.7/dist-packages (from gensim) (1.15.0)\n", | |
| "Requirement already satisfied: smart-open>=1.2.1 in /usr/local/lib/python3.7/dist-packages (from gensim) (5.1.0)\n", | |
| "Requirement already satisfied: scipy>=0.18.1 in /usr/local/lib/python3.7/dist-packages (from gensim) (1.4.1)\n", | |
| "Requirement already satisfied: numpy>=1.11.3 in /usr/local/lib/python3.7/dist-packages (from gensim) (1.19.5)\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "iEmTRgIY8sUy" | |
| }, | |
| "source": [ | |
| "#Second, train the word embedding model on the sequential list.\n", | |
| "import gensim\n", | |
| "from gensim.models import Word2Vec\n", | |
| "model = Word2Vec(sentences=nouns, size=200, window=4, min_count=1, workers=4) # Build custom Word2Vec" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "vI2p9Dwc5YxH", | |
| "outputId": "fb37e7d0-5bbb-4dad-a82b-d94d2efe2075" | |
| }, | |
| "source": [ | |
| "print(\"Total number of unique words loaded in Model : \", len(model.wv.vocab))" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "stream", | |
| "text": [ | |
| "Total number of unique words loaded in Model : 6294\n" | |
| ], | |
| "name": "stdout" | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "KkfgURolB7YI", | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "outputId": "8eb38a7d-0f3e-49a4-d247-7a8cac62a2db" | |
| }, | |
| "source": [ | |
| "model.wv.get_vector('high school') # Vector" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/plain": [ | |
| "array([ 1.17968558e-03, -6.13471901e-04, -4.19424381e-04, 1.39982541e-04,\n", | |
| " -4.42270451e-04, -1.52804633e-03, 1.25812506e-03, 1.54283503e-03,\n", | |
| " 1.59823848e-03, 1.24964968e-03, 9.47103545e-04, -9.90002882e-04,\n", | |
| " -1.86733296e-03, 3.27405549e-04, 2.51950999e-03, -1.98870269e-03,\n", | |
| " 7.34310306e-04, 2.20810436e-03, 4.74596396e-04, -1.23826414e-03,\n", | |
| " -8.87183851e-05, 1.66137310e-04, -7.56677473e-04, -1.74678964e-04,\n", | |
| " -2.13049515e-03, 5.45639021e-04, 2.22702301e-03, 1.74860423e-03,\n", | |
| " 3.93779948e-04, -1.96447060e-03, -1.85374019e-03, 3.10962030e-04,\n", | |
| " 2.38917815e-03, -2.21882993e-03, -1.56347652e-03, -1.36123621e-03,\n", | |
| " 2.30838195e-03, -1.09612674e-03, -9.07394686e-04, -1.38799602e-03,\n", | |
| " 3.53713229e-04, -2.14561843e-03, 5.66787960e-04, 1.14087027e-03,\n", | |
| " 2.03479826e-03, -9.81438207e-04, -9.59684956e-04, 3.64889391e-04,\n", | |
| " 1.63393328e-03, -7.15667557e-05, -1.33770716e-03, -7.43036333e-04,\n", | |
| " -4.78012837e-04, 2.16797693e-04, 8.61037755e-04, 7.98562251e-04,\n", | |
| " -1.04792509e-03, 1.65433958e-05, -8.07212899e-04, 7.49367871e-04,\n", | |
| " 7.84540607e-04, 1.61209749e-03, -1.63092290e-03, 5.12847982e-05,\n", | |
| " 1.37357740e-03, 4.64355806e-04, -1.90611999e-03, 1.66872074e-03,\n", | |
| " -3.91371548e-04, 1.80622481e-03, -1.64107012e-03, 5.29031502e-04,\n", | |
| " -1.29659078e-03, -1.82525814e-03, -2.42277212e-03, 2.26137671e-03,\n", | |
| " 1.80116924e-03, -1.53841847e-03, 1.99726899e-03, 1.99278304e-03,\n", | |
| " -8.18951346e-04, -1.09196233e-03, -2.43699993e-03, -2.20595323e-03,\n", | |
| " 9.15682700e-04, 1.27136253e-03, -1.69626263e-04, -2.04451964e-03,\n", | |
| " -9.27103567e-04, -1.48528174e-03, -2.43788771e-03, -8.09699995e-05,\n", | |
| " 2.21311813e-03, 4.51469532e-04, -1.14419358e-03, -2.00388301e-03,\n", | |
| " -1.21369341e-03, -2.17930973e-03, -1.32655900e-04, 9.59691824e-04,\n", | |
| " -1.84139237e-03, -1.32134976e-03, -1.76852720e-03, 2.23613507e-03,\n", | |
| " 6.10724324e-04, 2.15015770e-03, -1.25185191e-03, 1.25170205e-04,\n", | |
| " -3.37855163e-05, 2.09294097e-03, 1.50106079e-03, -5.50618977e-04,\n", | |
| " 1.80608197e-03, -4.11166693e-04, -3.74526920e-04, -1.22833892e-03,\n", | |
| " -7.39002251e-04, 6.85862673e-04, 2.17369245e-03, 2.13749800e-03,\n", | |
| " -7.88757927e-04, 1.34563039e-03, -1.06357526e-04, 2.50658533e-03,\n", | |
| " 2.98541618e-05, 1.38815865e-03, -1.99433998e-04, 8.17579159e-04,\n", | |
| " 1.57528801e-03, 4.39519848e-04, -1.16302061e-03, 1.60360488e-03,\n", | |
| " 1.33904652e-03, 1.74749049e-03, 1.84504373e-03, 1.09950826e-03,\n", | |
| " -1.72325678e-03, 2.12626455e-05, -2.51270784e-03, -1.47363357e-03,\n", | |
| " -1.90800882e-03, 2.15069368e-03, 2.01108516e-03, 4.99026610e-05,\n", | |
| " -8.35250714e-04, -6.45386055e-04, 1.71807176e-03, -1.70551974e-03,\n", | |
| " -1.84347655e-03, 2.41189636e-03, -1.92627043e-03, 1.93681498e-03,\n", | |
| " -8.95186677e-04, -2.34524673e-03, -1.75066048e-03, -1.70948252e-03,\n", | |
| " -1.83230592e-03, -2.10754792e-04, -1.77836919e-03, 6.12430507e-04,\n", | |
| " 8.99297534e-04, 6.91044494e-04, 7.51300773e-04, 2.17594858e-03,\n", | |
| " 1.43120182e-03, -2.40609189e-03, -2.11149664e-03, -2.48488178e-03,\n", | |
| " 1.31531223e-03, 2.14244632e-04, 1.07512774e-03, -1.72493677e-03,\n", | |
| " 1.64733711e-03, 1.50151900e-03, -1.36632676e-04, -1.77843112e-03,\n", | |
| " 2.10548285e-03, -1.83828652e-03, -3.60849022e-04, -2.72592355e-04,\n", | |
| " -1.65132689e-03, -2.99927517e-04, -6.24824781e-04, -2.01558927e-03,\n", | |
| " -2.20559700e-03, 1.95807894e-03, 4.36647999e-04, -1.37910817e-03,\n", | |
| " 2.04229727e-04, -5.88736846e-04, 2.35435757e-04, 1.36909052e-03,\n", | |
| " 7.57168396e-04, 7.54014764e-04, -6.29863876e-04, -1.53347279e-03,\n", | |
| " -3.99034092e-04, 1.73364286e-04, 3.82183935e-04, -5.87538016e-05],\n", | |
| " dtype=float32)" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 100 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "colab": { | |
| "base_uri": "https://localhost:8080/" | |
| }, | |
| "id": "hJ83So9l5ckG", | |
| "outputId": "7b1baa67-ac5d-4ae9-99b7-d1a53659caef" | |
| }, | |
| "source": [ | |
| "model.wv.most_similar(\"sports extracurriculars\", topn=10) # Distance based search between vectors" | |
| ], | |
| "execution_count": null, | |
| "outputs": [ | |
| { | |
| "output_type": "execute_result", | |
| "data": { | |
| "text/plain": [ | |
| "[('point view', 0.23088644444942474),\n", | |
| " ('son', 0.2303035408258438),\n", | |
| " ('scenes', 0.22999241948127747),\n", | |
| " ('quest', 0.22275203466415405),\n", | |
| " ('high school student teacher', 0.2195260226726532),\n", | |
| " ('ball', 0.2164124697446823),\n", | |
| " ('eureka', 0.21442648768424988),\n", | |
| " ('imagination quiet solitude cellar', 0.21432223916053772),\n", | |
| " ('young players', 0.2112780213356018),\n", | |
| " ('alarm', 0.20973749458789825)]" | |
| ] | |
| }, | |
| "metadata": { | |
| "tags": [] | |
| }, | |
| "execution_count": 106 | |
| } | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "zYVZqmcGggXi" | |
| }, | |
| "source": [ | |
| "**BONUS STEP**\n", | |
| "Use the word-embedding model to promote/demote suggestions from the N-gram model." | |
| ] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "43aYPe9YdV3G" | |
| }, | |
| "source": [ | |
| "## **I didn't have the time to figure out this problem**\n", | |
| "I feel a vector based kNN search would do the trick " | |
| ] | |
| }, | |
| { | |
| "cell_type": "code", | |
| "metadata": { | |
| "id": "d8Z3LXoNY_uh" | |
| }, | |
| "source": [ | |
| "#This is a bonus step\n", | |
| "#Using the word embedding model, filter out the distant/irrelevant suggestions from the N-gram next phrase prediction model.\n", | |
| "\n", | |
| "#HINT-> extract the noun using the POS tag from each of the next phrase predictions (in the ngram model)\n", | |
| "#Then find the distance of the NOUNS from the current NOUN in the word embedding model.\n", | |
| "#Rank the nouns in terms of distance from each other.\n", | |
| "#Retain only the top 5 results.\n", | |
| "\n" | |
| ], | |
| "execution_count": null, | |
| "outputs": [] | |
| }, | |
| { | |
| "cell_type": "markdown", | |
| "metadata": { | |
| "id": "jO-m0i19gxI2" | |
| }, | |
| "source": [ | |
| "**FINAL STEP** Submit the notebook to textifyai@gmail.com with your full name at the top of the document." | |
| ] | |
| } | |
| ] | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment