Toggle

backend/app.py

@@ -1,8 +1,10 @@
 import os
-from flask import Flask, render_template, request
+from flask import Flask, render_template, request, jsonify
 from flask_cors import CORS
 import pandas as pd
 import html
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.metrics.pairwise import cosine_similarity
 
 app = Flask(__name__)
 CORS(app)
@@ -11,9 +13,43 @@
 csv_file_path = 'data/semicleanedbg.csv'
 data_df = pd.read_csv(csv_file_path)
 
+# for cosine sim
+data_df['processed_description'] = data_df['description'].apply(lambda x: x.lower() if isinstance(x, str) else '')
+tfidf_vectorizer = TfidfVectorizer()
+tfidf_matrix = tfidf_vectorizer.fit_transform(data_df['processed_description'].values.astype('U'))  # 'U' for Unicode
+
+
+
+@app.route("/")
+def home():
+    return render_template('base.html', title="sample html")
+
+#for the suggestions box pls do not touch; it checks all the games for games with the query in it (not accounting for typos)
+# this then returns the top 5 matching titles
+@app.route("/suggestions")
+def suggestions():
+    query_partial = request.args.get("query", "")
+    if not query_partial:
+        return jsonify([])
+    matching_games = data_df[data_df['name'].str.contains(query_partial, case = False)]['name'].unique().tolist()
+    suggestions = matching_games[:5]
+    return jsonify(suggestions)
+
+
+@app.route("/games")
+def search():
+    text = request.args.get("title")
+    min_age = request.args.get("min_age", type=int)
+    min_players = request.args.get("min_players", type=int)
+    max_players = request.args.get("max_players", type=int)
+    category = request.args.get("category")
+    mode = request.args.get("mode")
+
+    if mode == 'recommendation':
+        matches = recommendation_search(text)
+    else:
+        matches = matches = data_df[data_df['name'].str.lower().str.contains(text.lower())]
 
-def csv_search(query, min_age, min_players, max_players, category):
-    matches = data_df[data_df['name'].str.lower().str.contains(query.lower())]
     if min_age is not None:
         matches = matches[matches['minage'] >= min_age]
     if min_players is not None:
@@ -22,25 +58,30 @@ def csv_search(query, min_age, min_players, max_players, category):
         matches = matches[matches['maxplayers'] <= max_players]
     if category:
         matches = matches[matches['boardgamecategory'].str.contains(category, case=False, na=False)]
-
     matches_filtered = matches[['name', 'description', 'average', 'objectid']] 
     matches_filtered['name'] = matches_filtered['name'].apply(html.unescape)
     matches_filtered['description'] = matches_filtered['description'].apply(html.unescape)
     matches_filtered_json = matches_filtered.to_json(orient='records')
+
     return matches_filtered_json
 
-@app.route("/")
-def home():
-    return render_template('base.html', title="sample html")
+def recommendation_search(query):
+    if query:
+        #some preprocessing: 1- fetch the game 2- get its processed desc
+        game_row = data_df[data_df['name'].str.lower() == query.lower()].iloc[0]
+        game_description = game_row['processed_description']
+        #cos sim
+        query_vector = tfidf_vectorizer.transform([game_description])
+        cosine_similarities = cosine_similarity(query_vector, tfidf_matrix).flatten()
+        #get the top 1000       
+        similar_indices = cosine_similarities.argsort()[-(1000+1):][::-1]
+    # game itself is the most similar one, so we exclude it if so.
+    similar_indices = similar_indices[1:]
 
-@app.route("/games")
-def episodes_search():
-    text = request.args.get("title")
-    min_age = request.args.get("min_age", type=int)
-    min_players = request.args.get("min_players", type=int)
-    max_players = request.args.get("max_players", type=int)
-    category = request.args.get("category")
-    return csv_search(text, min_age, min_players, max_players, category)
+    # Fetch the details of the top N similar games then return
+    similar_games = data_df.iloc[similar_indices]
+    return similar_games
+
 
 @app.route("/about/<game_id>")
 def about(game_id):