Mapping uri to words

mapping.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import difflib\n",
    "import re\n",
    "from nltk import pos_tag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df = pd.read_json('./data_v7.json')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['_id', 'corrected_question', 'sparql_query', 'sparql_template_id',\n",
       "       'verbalized_question'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Name the scientist whose supervisor was Ernest Rutherford and had a doctoral students named Charles Drummond Ellis?\n",
      "\n",
      "SELECT DISTINCT ?uri WHERE {?uri <http://dbpedia.org/ontology/doctoralAdvisor> <http://dbpedia.org/resource/Ernest_Rutherford> . ?uri <http://dbpedia.org/property/doctoralStudents> <http://dbpedia.org/resource/Charles_Drummond_Ellis>  . ?uri <https://www.w3.org/1999/02/22-rdf-syntax-ns#type> <http://dbpedia.org/ontology/Scientist>}\n",
      "\n",
      "What is the <scientist> whose <supervisor> is <Ernest Rutherford> and <doctoral students> is <Charles Drummond Ellis>?\n",
      "\n",
      "[('Name', 'NN'), ('the', 'DT'), ('scientist', 'NN'), ('whose', 'WP$'), ('supervisor', 'NN'), ('was', 'VBD'), ('Ernest', 'NNP'), ('Rutherford', 'NNP'), ('and', 'CC'), ('had', 'VBD'), ('a', 'DT'), ('doctoral', 'JJ'), ('students', 'NNS'), ('named', 'VBN'), ('Charles', 'NNP'), ('Drummond', 'NNP'), ('Ellis?', 'NNP')]\n"
     ]
    }
   ],
   "source": [
    "i = 10\n",
    "print(df.loc[i][1])\n",
    "print()\n",
    "print(df.loc[i][2])\n",
    "print()\n",
    "print(df.loc[i][4])\n",
    "print()\n",
    "print(pos_tag(df.loc[i][1].split()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "uris = re.findall(r'http://dbpedia.org/[^>]*', df.loc[i][2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['http://dbpedia.org/ontology/doctoralAdvisor',\n",
       " 'http://dbpedia.org/resource/Ernest_Rutherford',\n",
       " 'http://dbpedia.org/property/doctoralStudents',\n",
       " 'http://dbpedia.org/resource/Charles_Drummond_Ellis',\n",
       " 'http://dbpedia.org/ontology/Scientist']"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "uris"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "l = re.findall(r'\\S+[^?]', df.loc[i][1])\n",
    "l = [x.strip().lower() for x in l]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_dict = {}\n",
    "for uri in uris:\n",
    "    s = uri.split('/')\n",
    "    label, entity = s[-2], s[-1]\n",
    "    if label == 'resource':\n",
    "        entity = re.sub('_', ' ', entity).lower()\n",
    "        start = 1000\n",
    "        build = ''\n",
    "        for match in difflib.get_close_matches(entity, l, 5, 0):\n",
    "            if match in entity:\n",
    "                build += ' ' + match\n",
    "                temp = df.loc[i][1].lower().find(match)\n",
    "                if temp < start:\n",
    "                    start = temp\n",
    "            else:\n",
    "                break\n",
    "        build = build.strip()\n",
    "        length = len(build)\n",
    "        end = start + length - 1\n",
    "        final_dict.update({uri: [start, end]})\n",
    "    else:\n",
    "        continue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'http://dbpedia.org/resource/Charles_Drummond_Ellis': [92, 113],\n",
       " 'http://dbpedia.org/resource/Ernest_Rutherford': [40, 56]}"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_dict"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}