A Hybrid AI-Based Educational Application Recommendation System Using Content Filtering and User Preference Modeling

Abstract

In the rapidly evolving digital learning ecosystem, users are often overwhelmed by the
vast number of educational applications available across multiple domains. Identifying
the most relevant and effective learning tools tailored to individual needs remains a
significant challenge. This project presents a Hybrid AI-Based Educational
Application Recommendation System designed to deliver personalized suggestions
using a combination of content-based filtering and user preference modeling. The system
is developed using the Django web framework and integrates machine learning
techniques to enhance user experience. It utilizes content-based filtering by analyzing
application metadata such as category and description using Term Frequency-Inverse
Document Frequency (TF-IDF) and cosine similarity. This enables the system to
recommend applications that are similar in content to those previously explored by the
user. In addition, the system incorporates user preference-based personalization, where
user attributes such as interested categories, education level, job role, and location are
considered. A scoring mechanism evaluates how well each application aligns with the
user’s profile, ensuring highly relevant recommendations. To improve diversity and avoid
recommendation stagnation, controlled randomness is introduced in the ranking process.
The system also integrates a crowdsourcing module, allowing users to provide ratings,
reviews, complexity levels, and tags for applications. This user-generated data
contributes to improving recommendation accuracy over time. Furthermore, a deep
learning model using TensorFlow is proposed as an extension to predict user ratings
based on combined user and application features, enabling future scalability and
intelligence. The platform supports essential functionalities such as user authentication,
search capabilities, detailed application views, and recommendation dashboards. It
ensures a seamless and interactive user experience while maintaining modularity and
scalability in design