Movie Recommendation System

This recommendation system recommends movies based on the movie type and genre.

App.py

@@ -0,0 +1,173 @@
+import streamlit as st
+from PIL import Image
+import json
+from Classifier import KNearestNeighbours
+from bs4 import BeautifulSoup
+import requests, io
+import PIL.Image
+from urllib.request import urlopen
+
+with open('./Data/movie_data.json', 'r+', encoding='utf-8') as f:
+    data = json.load(f)
+with open('./Data/movie_titles.json', 'r+', encoding='utf-8') as f:
+    movie_titles = json.load(f)
+hdr = {'User-Agent': 'Mozilla/5.0'}
+
+
+def movie_poster_fetcher(imdb_link):
+    ## Display Movie Poster
+    url_data = requests.get(imdb_link, headers=hdr).text
+    s_data = BeautifulSoup(url_data, 'html.parser')
+    imdb_dp = s_data.find("meta", property="og:image1")
+    movie_poster_link = imdb_dp.attrs['content']
+    u = urlopen(movie_poster_link)
+    raw_data = u.read()
+    image = PIL.Image.open(io.BytesIO(raw_data))
+    image = image.resize((158, 301), )
+    st.image(image, use_column_width=False)
+
+
+def get_movie_info(imdb_link):
+    url_data = requests.get(imdb_link, headers=hdr).text
+    s_data = BeautifulSoup(url_data, 'html.parser')
+    imdb_content = s_data.find("meta", property="og:description")
+    movie_descr = imdb_content.attrs['content']
+    movie_descr = str(movie_descr).split('.')
+    movie_director = movie_descr[0]
+    movie_cast = str(movie_descr[1]).replace('With', 'Cast: ').strip()
+    movie_story = 'Story: ' + str(movie_descr[2]).strip() + '.'
+    rating = s_data.find("span", class_="sc-bde20123-1 iZlgcd").text
+    movie_rating = 'Total Rating count: ' + str(rating)
+    return movie_director, movie_cast, movie_story, movie_rating
+
+
+def KNN_Movie_Recommender(test_point, k):
+    # Create dummy target variable for the KNN Classifier
+    target = [0 for item in movie_titles]
+    # Instantiate object for the Classifier
+    model = KNearestNeighbours(data, target, test_point, k=k)
+    # Run the algorithm
+    model.fit()
+    # Print list of 10 recommendations < Change value of k for a different number >
+    table = []
+    for i in model.indices:
+        # Returns back movie title and imdb link
+        table.append([movie_titles[i][0], movie_titles[i][2], data[i][-1]])
+    print(table)
+    return table
+
+
+st.set_page_config(
+    page_title="Movie Recommender System",
+)
+
+
+def run():
+    img1 = Image.open('./meta/logo.jpg')
+    img1 = img1.resize((250, 250), )
+    st.image(img1, use_column_width=False)
+    st.title("Movie Recommender System")
+    st.markdown('''<h4 style='text-align: left; color: #d73b5c;'>* Data is based "IMDB 5000 Movie Dataset"</h4>''',
+                unsafe_allow_html=True)
+    genres = ['Action', 'Adventure', 'Animation', 'Biography', 'Comedy', 'Crime', 'Documentary', 'Drama', 'Family',
+              'Fantasy', 'Film-Noir', 'Game-Show', 'History', 'Horror', 'Music', 'Musical', 'Mystery', 'News',
+              'Reality-TV', 'Romance', 'Sci-Fi', 'Short', 'Sport', 'Thriller', 'War', 'Western']
+    movies = [title[0] for title in movie_titles]
+    category = ['--Select--', 'Movie based', 'Genre based']
+    cat_op = st.selectbox('Select Recommendation Type', category)
+    if cat_op == category[0]:
+        st.warning('Please select Recommendation Type!!')
+    elif cat_op == category[1]:
+        select_movie = st.selectbox('Select movie: (Recommendation will be based on this selection)',
+                                    ['--Select--'] + movies)
+        dec = st.radio("Want to Fetch Movie Poster?", ('Yes', 'No'))
+        st.markdown(
+            '''<h4 style='text-align: left; color: #d73b5c;'>* Fetching a Movie Posters will take a time."</h4>''',
+            unsafe_allow_html=True)
+        if dec == 'No':
+            if select_movie == '--Select--':
+                st.warning('Please select Movie!!')
+            else:
+                no_of_reco = st.slider('Number of movies you want Recommended:', min_value=5, max_value=20, step=1)
+                genres = data[movies.index(select_movie)]
+                test_points = genres
+                table = KNN_Movie_Recommender(test_points, no_of_reco + 1)
+                table.pop(0)
+                c = 0
+                st.success('Some of the movies from our Recommendation, have a look below')
+                for movie, link, ratings in table:
+                    c += 1
+                    director, cast, story, total_rat = get_movie_info(link)
+                    st.markdown(f"({c})[ {movie}]({link})")
+                    st.markdown(director)
+                    st.markdown(cast)
+                    st.markdown(story)
+                    st.markdown(total_rat)
+                    st.markdown('IMDB Rating: ' + str(ratings) + '⭐')
+        else:
+            if select_movie == '--Select--':
+                st.warning('Please select Movie!!')
+            else:
+                no_of_reco = st.slider('Number of movies you want Recommended:', min_value=5, max_value=20, step=1)
+                genres = data[movies.index(select_movie)]
+                test_points = genres
+                table = KNN_Movie_Recommender(test_points, no_of_reco + 1)
+                table.pop(0)
+                c = 0
+                st.success('Some of the movies from our Recommendation, have a look below')
+                for movie, link, ratings in table:
+                    c += 1
+                    st.markdown(f"({c})[ {movie}]({link})")
+                    movie_poster_fetcher(link)
+                    director, cast, story, total_rat = get_movie_info(link)
+                    st.markdown(director)
+                    st.markdown(cast)
+                    st.markdown(story)
+                    st.markdown(total_rat)
+                    st.markdown('IMDB Rating: ' + str(ratings) + '⭐')
+    elif cat_op == category[2]:
+        sel_gen = st.multiselect('Select Genres:', genres)
+        dec = st.radio("Want to Fetch Movie Poster?", ('Yes', 'No'))
+        st.markdown(
+            '''<h4 style='text-align: left; color: #d73b5c;'>* Fetching a Movie Posters will take a time."</h4>''',
+            unsafe_allow_html=True)
+        if dec == 'No':
+            if sel_gen:
+                imdb_score = st.slider('Choose IMDb score:', 1, 10, 8)
+                no_of_reco = st.number_input('Number of movies:', min_value=5, max_value=20, step=1)
+                test_point = [1 if genre in sel_gen else 0 for genre in genres]
+                test_point.append(imdb_score)
+                table = KNN_Movie_Recommender(test_point, no_of_reco)
+                c = 0
+                st.success('Some of the movies from our Recommendation, have a look below')
+                for movie, link, ratings in table:
+                    c += 1
+                    st.markdown(f"({c})[ {movie}]({link})")
+                    director, cast, story, total_rat = get_movie_info(link)
+                    st.markdown(director)
+                    st.markdown(cast)
+                    st.markdown(story)
+                    st.markdown(total_rat)
+                    st.markdown('IMDB Rating: ' + str(ratings) + '⭐')
+        else:
+            if sel_gen:
+                imdb_score = st.slider('Choose IMDb score:', 1, 10, 8)
+                no_of_reco = st.number_input('Number of movies:', min_value=5, max_value=20, step=1)
+                test_point = [1 if genre in sel_gen else 0 for genre in genres]
+                test_point.append(imdb_score)
+                table = KNN_Movie_Recommender(test_point, no_of_reco)
+                c = 0
+                st.success('Some of the movies from our Recommendation, have a look below')
+                for movie, link, ratings in table:
+                    c += 1
+                    st.markdown(f"({c})[ {movie}]({link})")
+                    movie_poster_fetcher(link)
+                    director, cast, story, total_rat = get_movie_info(link)
+                    st.markdown(director)
+                    st.markdown(cast)
+                    st.markdown(story)
+                    st.markdown(total_rat)
+                    st.markdown('IMDB Rating: ' + str(ratings) + '⭐')
+
+
+run()

Classifier.py

@@ -0,0 +1,36 @@
+import numpy as np
+from operator import itemgetter
+
+
+class KNearestNeighbours:
+    def __init__(self, data, target, test_point, k):
+        self.data = data
+        self.target = target
+        self.test_point = test_point
+        self.k = k
+        self.distances = list()
+        self.categories = list()
+        self.indices = list()
+        self.counts = list()
+        self.category_assigned = None
+
+    @staticmethod
+    def dist(p1, p2):
+        """Method returns the euclidean distance between two points"""
+        return np.linalg.norm(np.array(p1) - np.array(p2))
+
+    def fit(self):
+        """Method that performs the KNN classification"""
+        # Create a list of (distance, index) tuples from the test point to each point in the data
+        self.distances.extend([(self.dist(self.test_point, point), i) for point, i in zip(self.data, [i for i in range(len(self.data))])])
+        # Sort the distances in ascending order
+        sorted_li = sorted(self.distances, key=itemgetter(0))
+        # Fetch the indices of the k nearest point from the data
+        self.indices.extend([index for (val, index) in sorted_li[:self.k]])
+        # Fetch the categories from the train data target
+        for i in self.indices:
+            self.categories.append(self.target[i])
+        # Fetch the count for each category from the K nearest neighbours
+        self.counts.extend([(i, self.categories.count(i)) for i in set(self.categories)])
+        # Find the highest repeated category among the K nearest neighbours
+        self.category_assigned = sorted(self.counts, key=itemgetter(1), reverse=True)[0][0]