Create collaborative_filtering.py

collaborative_filtering.py

@@ -0,0 +1,20 @@
+from surprise import Dataset, Reader, SVD
+from surprise.model_selection import train_test_split, accuracy
+
+# Load data from database
+data = cursor.execute('SELECT user_id, item_id, interaction_type FROM user_interactions').fetchall()
+
+# Convert to Surprise dataset
+reader = Reader(rating_scale=(1, 5))
+dataset = Dataset.load_from_df(pd.DataFrame(data, columns=['user_id', 'item_id', 'interaction_type']), reader)
+
+# Train-test split
+trainset, testset = train_test_split(dataset, test_size=0.25)
+
+# Train model
+algo = SVD()
+algo.fit(trainset)
+
+# Test model
+predictions = algo.test(testset)
+accuracy.rmse(predictions)