How nehathakur4/price-negotiator-ecommerce-chatbot-system works
Project Type and Purpose
• This is a full-stack e-commerce chatbot system that enables customers to negotiate product prices in natural language.
• Key features: user onboarding (signup/login), product browsing, NLP-driven chat interface, machine-learning-based price prediction, reinforcement-learning negotiation engine, and payment processing.
• Built with Python (Flask backend), HTML/CSS frontend, TensorFlow/Keras, NLTK, Scikit-learn, and an Anaconda-managed environment.
File Structure & README
• Top level:
• Price-Negotiator-Ecommerce-Chatbot-System.pdf (project docs)
• README.md (project overview, modules, screenshots)
• screenshots/ (UI flow images)
• “chatbot_code.py” (core logic orchestrating NLP, ML, RL)
• “negotiation_data.csv” (training dataset)
• Presence of Flask implies an HTTP API layer; HTML/CSS implies a client served by Flask; payment page indicates an external payment gateway integration.
• Likely a simple relational or key-value store (not explicitly shown) for users, products, and negotiation state.
Extract Insights from README
• Modules:
a. NLP Module (intent & entity extraction)
b. Price Prediction Module (historical-data ML model)
c. Negotiation Engine (reinforcement learning)
• UI Steps (screenshots): signup/login → product list → chat interface → checkout/payment → success.
• External dependencies: payment gateway; uses Flask for backend and NLTK/TensorFlow for AI.
Main Components & Interactions
• Frontend (Browser UI)
• Signup/Login Page
• Product Listing & Selection
• Chat Interface (negotiation)
• Payment & Confirmation Pages
• Backend (Flask App)
• REST/WS endpoint for authentication
• Product service (fetch product data)
• Chatbot controller that routes messages to NLP Module → Price Prediction → Negotiation Engine → response.
• Payment integration module calling external payment gateway API.
• Data Storage
• User Database (accounts, sessions)
• Product Database (pricing, inventory)
• Negotiation Data Store (conversation logs, training data in CSV)
• ML Model Store (serialized TensorFlow/Keras models)
• External Services
• Payment Gateway (e.g., Stripe/PayPal)
• (Optional) Third-party analytics or logging
Architectural Patterns & Design Principles
• Layered Architecture: Presentation (frontend) → Application (Flask controllers) → Business Logic (NLP/ML/Negotiation) → Data
• Loose coupling: each module exposes a clear interface
• Data flow: user message → NLP → intent/entities → price prediction → negotiation decision → JSON response → frontend display