Chess may seem simple at first because the rules are clear, but it is actually one of the most complex games. With almost unlimited possible moves, it is nearly impossible to master all combinations, which makes learning chess very challenging. To help players learn and improve more effectively, a tool was developed as part of a thesis project. It is based on a personalized approach to training with chess puzzles. These puzzles are created by analyzing the user's games from Chess.com and Lichess. Instead of using general exercises, the application offers puzzles based on the actual mistakes made by the player and focuses on their tactical weaknesses. The system automatically finds key positions, filters out unsuitable puzzles, and evaluates them with a machine learning model to show only the most relevant ones. In addition to tactical puzzles, the app also offers analysis of opening mistakes and an overview of how successfully the user recognizes tactical motifs in their own games. Opening mistakes can be replayed, and tactical motifs can be practiced through a link to the Lichess platform. The goal of the application is to help each player discover their own weaknesses and improve them through targeted practice.
|
