Sentiment Analysis API

Overview

This repository provides a basic API for sentiment analysis, utilizing the OpenAI API for processing text inputs. The API categorizes sentiment as positive, negative, or neutral.

Key Features

• Sentiment Analysis: Analyzes and categorizes text sentiment.
• OpenAI API Integration: Uses OpenAI API for text analysis.
• User Authentication: Users must be logged in to use the API.
• Customizable Categories: Users can customize sentiment categories. Defaults are positive, negative, and neutral.
• Rate Limiter: Limits usage to 100 requests per 15 minutes.
• Daily Usage Limit: Each user has 150 requests per day, renewed nightly at 00:00.
• Redis Integration: Utilizes Redis for efficient in-memory data storage and management.

Used Technologies

• express: Web framework for Node.js.
• mongoose: MongoDB object modeling tool.
• jsonwebtoken: JSON Web Token implementation.
• bcryptjs: Library to hash passwords.
• express-validator: Middleware for validating and sanitizing user inputs
• dotenv: Module to load environment variables.
• express-rate-limit: Middleware to limit repeated requests.
• node-cron:Task scheduler for Node.js.
• nodemon: Utility to monitor for changes in Node.js applications.
• openai: For sentiment analysis.
• redis: In-memory data structure store.

Endpoints

• POST /register

http://localhost:3000/register

Headers:

Content-Type: application/json

Request:

{
    "name": "ali",
    "email":"a@gmail.com",
    "password":"123456"
}

Response:

{
  "success": true,
  "token": ""
}

---

• POST /login

http://localhost:3000/login

Headers:

Content-Type: application/json

Request:

{
    "email":"a@gmail.com",
    "password":"123456"
}

Response:

{
  "success": true,
  "token": ""
}

---

• POST /analyze

http://localhost:3000/analyze

Headers:

Content-Type: application/json
Authorization: Bearer your_access_token_here

Request:

{
    "text": "The customer service was not good enough, and I am very disappointed with the experience."
}

Response:

{
  "success": true,
  "data": "Negative"
}

---

• POST /custom-sentiment-categories

http://localhost:3000/custom-sentiment-categories

Headers:

Content-Type: application/json
Authorization: Bearer your_access_token_here

Request:

{
  "categories": ["Happy", "Sad", "Angry"]
}

Response:

{
  "success": true,
  "data": [
    "Happy",
    "Sad",
    "Angry"
  ]
}

---

• GET /profile

http://localhost:3000/profile

Headers:

Content-Type: application/json
Authorization: Bearer your_access_token_here

Response:

{
  "success": true,
  "data": {
    "_id": "6666059d80df1c6f48e43c1f",
    "name": "ali",
    "email": "a@gmail.com",
    "customSentimentCategories": [
      "Happy",
      "Sad",
      "Angry"
    ],
    "requestCount": 1,
    "requestLimit": 2,
    "createdAt": "2024-06-09T19:42:21.608Z",
    "__v": 1
  }
}

---

Setup

1.Clone the Repo:

git clone https://github.com/carpodok/sentiment-analysis-api.git

2. Install Dependencies:

npm install

3. Configure Environment Variables

Create a .envfile in the root directory and add the following variables:

PORT=3000
OPENAI_API_KEY=your_api_key
MONGO_URI="your_mongo_uri"
JWT_SECRET=jwt_secret
REDIS_URL=redis://127.0.0.1:6379

4. Start the server:

nem start

For development purpose with live reload, use this:

npm run dev

License

This project is lisenced under the MIT License.