RAG-product-recommendation

This repository uses the OpenAI API to connect to the Elasticsearch search engine and provide product recommendations to users.

Purpose

Allow users to input their questions on the front end, search the database based on the users' questions, and respond with suitable products for the users.

Preparation

1. Install exe Elasticsearch
2. Install python package: pip install flask
3. Start Elasticsearch
4. Run app.py
5. Use python -m http.server 20000 to start index.html.

Code Description

Elasticsearch

• elast.py:
  Read SalePage.csv and input the product names and product descriptions into Elasticsearch.
• elastSh.py:
  Query what values are in Elasticsearch.
• elashDel.py:
  Delete all products in Elasticsearch.

Produce_html

• generate_products_html.py:
  Generate the front-end homepage, search button, and product page.
• SalePage.7z:
  It contains product data. Please unzip it and use the password to unlock.

Frontend_Backend

• api_key.txt:
  Store the OpenAI API key, please use your own.
• app.py:
  Use a Flask-written backend to connect to Elasticsearch, send the values returned by Elasticsearch to the OpenAI API for GPT-4, and finally send the values returned by GPT-4 to index.html.
• index.html:
  Front-end homepage
• products.html:
  Product page