Personalized News Recommendation System

Introduction

Problem Identification

With the massive amount of content that an average person consumes on a day to day basis, information overload is a concern. Users looking for customized contents are rising and people prefer to only hear and learn about things of their interest. This issue is been visible in media platforms specifically in terms of NEWS.

Solution

The proposed solution takes into account the users interactions such as viewing, skipping, liking, rating an article. It also has a methodology which takes into account the collaborative interactions of users in order to recommend the most suitable and relevant content. By web scraping existing popular news sites, we provide the user with the opportunity to see the news ranked according to their preference. A comprehensive solution has been developed to take in user input, web scrape, store the data in the database, carry on the necessary tasks, retrieve and amend data as needed, with the ultimate objective of recommending the most suitable news for the user according to their predicted preference.

Features

• Content-Based and Collaborative Filtering
• User Interaction Tracking (likes, skips, ratings)
• Multi-user Management
• Hybrid Recommendation System
• Real-time web scraping
• Administration access to update keywords, add or remove articles

Usage

1. Log in or register as a new user.
2. Interact with articles to personalize recommendations.

Technologies Used

• Java: Core logic
• MySQL: Database