File size: 3,213 Bytes
5371986 2f249ea 5371986 2f249ea 5371986 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 | import pandas as pd
import numpy as np
import sklearn
from sklearn.metrics.pairwise import cosine_similarity
# Load movie metadata into a pandas DataFrame
movie_data = pd.read_csv('movies_metadata.csv', usecols=['title', 'budget', 'revenue', 'vote_count', 'vote_average'])
# Load movie ratings into a pandas DataFrame
ratings_data = pd.read_csv('ratings_small.csv')
# Collaborative Filtering
def collaborative_filtering(user_id, num_recommendations=5):
user_ratings = ratings_data[ratings_data['userId'] == user_id]
user_ratings = user_ratings.merge(movie_data[['title']], left_on='movieId', right_index=True)
user_ratings = user_ratings.drop_duplicates(subset=['title'], keep='first')
# Get the mean ratings for each movie
movie_ratings_mean = ratings_data.groupby('movieId')['rating'].mean().reset_index()
# Merge with movie data to get movie titles
movie_ratings_mean = movie_ratings_mean.merge(movie_data[['title']], left_on='movieId', right_index=True)
movie_ratings_mean = movie_ratings_mean.drop_duplicates(subset=['title'], keep='first')
# Sort movies by mean rating in descending order
top_movies = movie_ratings_mean.sort_values(by='rating', ascending=False)['title'][:num_recommendations]
return top_movies.tolist()
# Content-Based Filtering
def content_based_filtering(movie_title, num_recommendations=5):
# Drop duplicates and missing values based on 'title' column
movie_data_unique = movie_data.drop_duplicates(subset='title', keep='first').dropna(subset=['budget', 'revenue', 'vote_count', 'vote_average'])
# Calculate similarity scores between input movie and other movies based on selected features
selected_features = ['budget', 'revenue', 'vote_count', 'vote_average']
movie_features = movie_data_unique[selected_features]
similarity_scores = cosine_similarity(movie_features[movie_data_unique['title'] == movie_title], movie_features)
movie_data_unique['similarity'] = similarity_scores.flatten()
# Get the most similar movies
top_movies = movie_data_unique[movie_data_unique['title'] != movie_title].sort_values(by='similarity', ascending=False)['title'][:num_recommendations]
return top_movies.tolist()
# Hybrid Method
def hybrid_filtering(user_id, movie_title, num_recommendations=5):
collaborative_movies = collaborative_filtering(user_id)
content_based_movies = content_based_filtering(movie_title)
# Merge and deduplicate the recommendations
all_recommendations = collaborative_movies + content_based_movies
unique_recommendations = list(set(all_recommendations))[:num_recommendations]
return unique_recommendations
import gradio as gr
with gr.Blocks() as demo:
gr.Markdown("Flip text or image files using this demo.")
title="Movie Recommender System"
user_id = gr.Textbox(placeholder='user_id')
movie_title = gr.Textbox(placeholder='movie_title')
#input2 = gr.Textbox(placeholder='second film name...')
#input3 = gr.Textbox(placeholder='third film name...')
text_output=gr.Textbox( )
text_button = gr.Button("Submit")
text_button.click(hybrid_filtering, inputs= [ user_id,movie_title], outputs=text_output)
demo.launch() |