product_recommender.py

import pandas as pd
from sentence_transformers import SentenceTransformer
import faiss

class ProductRecommender:
    def __init__(self, product_data_path):
        try:
            # Attempt to load the product data CSV
            self.data = pd.read_csv(product_data_path)
            print("Product data loaded successfully.")
        except Exception as e:
            print(f"Error loading product data: {e}")
            self.data = pd.DataFrame()  # Create an empty DataFrame if loading fails
            return

        try:
            # Initialize the sentence transformer model
            self.model = SentenceTransformer('all-MiniLM-L6-v2')
            print("Model loaded successfully.")
        except Exception as e:
            print(f"Error loading SentenceTransformer model: {e}")
            self.model = None  # Set model to None if loading fails
            return

        try:
            # Check if 'product_description' column exists
            if 'product_description' not in self.data.columns:
                print("Error: 'product_description' column is missing in the data.")
                return

            # Generate embeddings for the product descriptions
            self.embeddings = self.model.encode(self.data['product_description'].tolist())
            print(f"Embeddings generated successfully. Shape: {self.embeddings.shape}")
        except Exception as e:
            print(f"Error generating embeddings: {e}")
            self.embeddings = None  # Set embeddings to None if generation fails
            return

        try:
            # Initialize FAISS index and add the embeddings
            self.index = faiss.IndexFlatL2(self.embeddings.shape[1])
            self.index.add(self.embeddings)
            print("FAISS index created and embeddings added.")
        except Exception as e:
            print(f"Error creating FAISS index or adding embeddings: {e}")
            self.index = None  # Set index to None if creation fails
            return

    def get_recommendations(self, query, top_n=5):
        if self.model is None or self.index is None:
            print("Error: Model or FAISS index not initialized. Cannot make recommendations.")
            return []

        try:
            # Generate the embedding for the query
            query_embedding = self.model.encode([query])
            print(f"Query embedding generated. Shape: {query_embedding.shape}")
        except Exception as e:
            print(f"Error generating query embedding: {e}")
            return []

        try:
            # Search for top_n recommendations
            distances, indices = self.index.search(query_embedding, top_n)
            recommendations = []
            for i in indices[0]:
                recommendations.append(self.data.iloc[i]['product_title'] + ": " + self.data.iloc[i]['product_description'])
            print(f"Recommendations generated successfully: {recommendations}")
            return recommendations
        except Exception as e:
            print(f"Error during recommendation search: {e}")
            return []