Social Network Website Using Link Prediction Models for Follower Suggestions

Project Introduction

We developed a social network website inspired by Instagram. The project includes a website for users and another for administrators using a NoSQL database. The user site provides all the basic features of a social network such as posting, liking, commenting on posts, as well as following, messaging, etc. between users. The admin site allows for statistical analysis and management of content such as posts, users, user reports, and more. In addition, using graph theory and link prediction algorithms, we employ four link prediction algorithms based on node similarity: Common Neighbor, Jaccard, Adamic-Adar, and Katz Index.

• This project is used to serve as a final term project for two courses in the Information Systems department at UIT: Web Application Development - IS207 (9.5/10) and Social Network - IS353 (9/10)

Team Members

ID	Name	Facebook	Contribution %
21522791	Nguyễn Hoàng Việt	Nguyễn Hoàng Việt	100
21520400	Nguyễn Hoàng Phúc	Nguyễn Hoàng Phúc	100

Technologies Used

• Frontend: React, Redux, Bootstrap
• Backend: NodeJS, ExpressJS, FastAPI
• Real-time: Socket.IO
• Database: MongoDB
• Graph Theory & Link Prediction: NetworkX

Database Schema

Main Features of the Website

---

Main Features for Users

• Login / Register (API Token)
• Create/Like/Comment/Report Posts
• Follow/Search/View Other Users' Profiles
• Real-time Notifications/Messaging/Calling
• Manage Own Profile

Main Features for Admins

• Dashboard
• Manage Posts
• Manage Users
• Manage Post Reports

Demo of Some Interfaces

User Interface

• Login

• Home Page

• Create Post

• Comment on Post

• Search and Explore

• Messaging

• Notifications and Profile

Admin Interface

• Dashboard

• Manage Posts

• Manage Users

• User Statistics and Send Email

• Manage Post Reports

Installation Guide

Prerequisites

• Node.js
• npm or yarn
• Python 3.x (for FastAPI and NetworkX)
• MongoDB (local or cloud instance)

Frontend Setup

1. Clone the repository and navigate to the client directory:

   git clone https://github.com/NHViet03/Web_Social_Network_with_Link_Prediction
   cd /Web_Social_Network_with_Link_Prediction/client
   

2. Install dependencies:

   npm install
   

3. Start the client frontend server:

   npm start
   

4. Navigate to the admin directory and install dependencies:

   cd /Web_Social_Network_with_Link_Prediction/admin
   npm install
   

5. Start the admin frontend server:

   npm start
   

Backend Setup

1. Navigate to the backend directory and Install dependencies:

   cd /Web_Social_Network_with_Link_Prediction
   npm install
   

2. Set up environment variables, create a .env file with:

   MONGODB_URL=your_mongodb_connection_string
   ACCESS_TOKEN_SECRET=your_access_key
   REFRESH_TOKEN_SECRET=your_refresh_key
   

3. Start the backend server:

   npm start
   

FastAPI Setup (for follower suggestion)

1. Install FastAPI, NetworkX and required libraries:

   pip install fastapi uvicorn networkx
   npm install
   

2. Navigate to the FastAPI directory:

   cd /Web_Social_Network_with_Link_Prediction/fastAPIServer
   

3. Start the FastAPI server::

   python -m uvicorn index:app --reload
   

Related Project

• Mobile-App_Social_Network

References

• MERN Stack - Build a social media app