Last active
January 16, 2024 00:56
-
-
Save meow-d/8e32ef0a843e284ff45d043f9844eeef to your computer and use it in GitHub Desktop.
easiest way to implement the recipe recommendation algorithm
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
import re | |
import pandas as pd | |
from sklearn.feature_extraction.text import TfidfVectorizer | |
from sklearn.metrics.pairwise import linear_kernel | |
# load dataset | |
dataset = pd.read_csv("recipe_nlg_lite/train.csv", sep=" ") | |
dataset["combined"] = dataset.apply( | |
lambda row: " ".join(row[["name", "description", "ner", "steps"]]), axis=1 | |
) | |
# alternative stopwords and preprocessing | |
# with open("stopwords.txt", "r") as file: | |
# stop_words = file.read().splitlines() | |
# def no_number_preprocessor(tokens): | |
# r = re.sub("(\d)+", "NUM", tokens.lower()) | |
# # This alternative just removes numbers: | |
# # r = re.sub('(\d)+', '', tokens.lower()) | |
# return r | |
# tfidf_desc = TfidfVectorizer(stop_words=stop_words, preprocessor=no_number_preprocessor) | |
# vectorize | |
tfidf_desc = TfidfVectorizer(stop_words="english") | |
tfidf_desc_matrix = tfidf_desc.fit_transform(dataset["combined"]) | |
# calculate similarity, using linear kernel or cosine similarity | |
sim_cosine_desc = linear_kernel(tfidf_desc_matrix, tfidf_desc_matrix) | |
# functions | |
indices = pd.Series(dataset.index, index=dataset["uid"]).drop_duplicates() | |
def get_recommendations(uid, similarity, num_recommend=10): | |
idx = indices[ | |
uid | |
] # Get the pairwsie similarity scores of all movies with that movie | |
sim_scores = list( | |
enumerate(similarity[idx]) | |
) # Sort the movies based on the similarity scores | |
sim_scores = sorted( | |
sim_scores, key=lambda x: x[1], reverse=True | |
) # Get the scores of the 10 most similar movies | |
top_similar = sim_scores[1 : num_recommend + 1] # Get the movie indices | |
recipe_indices = [i[0] for i in top_similar] | |
return dataset.loc[recipe_indices] | |
# actually use the system | |
a = get_recommendations( | |
"dab8b7d0-e0f6-4bb0-aed9-346e80dace1f", | |
similarity=sim_cosine_desc, | |
num_recommend=20, | |
) | |
# recommendation_based_on_ingr = get_recommendations( | |
# "dab8b7d0-e0f6-4bb0-aed9-346e80dace1f", | |
# cosine_sim=cosine_sim_ingr, | |
# num_recommend=20, | |
# ) | |
a.to_csv("output/description_cosine.csv") |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment