A Web-Based Recommender Application for Board Games Using OOP Design

Today, board games are widely used for educational purposes. Scope of board games used in education covers various fields such as history, strategy, economy, and military. In this paper, we present an Object-Oriented Programming (OOP) design and online implementation of a web-based recommender application for board games, employing three popular recommendation techniques. The proposed system employs item-based collaborative filtering, user-based collaborative filtering, and content-based recommendation. We design the system's architecture exploiting OOP concepts such as inheritance and polymorphism. By utilizing OOP concepts, the proposed system achieves modularity, extensibility, and maintainability, harnessing three recommendation techniques in a compact structure. To the best of our knowledge, such a modular design for the conventional recommendation techniques has not been presented previously. Additionally, we implemented a data preprocessing pipeline to ensure the system's readiness for online deployment with the board games dataset. This pipeline encompasses strategic data design, data cleaning, and transformative data handling procedures. As a result, our system handles information retrieval from a vast dataset comprising approximately 19 million ratings for over 20,000 games. By integrating three different methodologies in the system, we aim to present a variety of game recommendations for educators, but for also board game players as well.