Real-time Chat Application with Sentiment Analysis

A chat application built with ASP.NET Core and SignalR, featuring AI-powered sentiment analysis using Azure Language Service. The application analyzes the emotional tone of each message and displays sentiment scores in real-time.

Architecture

Backend (ASP.NET Core)

• Framework: .NET 9.0
• Real-time Communication: SignalR with Azure SignalR Service
• Database: Entity Framework Core with SQL Server
• AI Service: Azure Cognitive Services (Language service)
• API: RESTful endpoints for message retrieval

Frontend

• HTML5/CSS3/JavaScript: Vanilla JavaScript for lightweight performance
• SignalR Client: Microsoft SignalR JavaScript library
• Real-time Updates: Websocket connection for instant message delivery

API Endpoints

REST API

• GET /api/chat/messages - Retrieve chat history
• GET /api/chat/messages/{id} - Get specific message
• GET /api/chat/health - Health check endpoint

SignalR Hub

• Hub URL: /chathub

Key Backend Components

ChatHub.cs

SignalR hub that handles real-time message broadcasting. Receives messages from clients. Prforms sentiment analysis and saves all messages to database.

SentimentAnalysisService.cs

Integrates with Azure Language Service to analyze message sentiment. Returns scores for positive, negative and neutral sentiment.

ChatController.cs

REST API controller providing endpoints to retrieve messages.