Enhancing Early Childhood Literacy: A Technological Aid for Letter Learning

Children generally learn at a slower pace than adults, necessitating additional practice, especially during early education stages. Mastery of the alphabet is crucial as it forms the foundation of any language, making it imperative for children to learn it first. Each child possesses a distinct learning style, and the duration required to grasp new concepts can vary considerably among individuals. To address this, the Sinhala Alphabet Learning Tool has been developed, offering preschoolers an interactive platform that enhances engagement and facilitates their learning experience. This research employs the K-Nearest Neighbors algorithm combined with image processing techniques to accurately predict letters input by preschoolers. A substantial dataset comprising 56,000 entries was meticulously collected directly from preschool participants. The data were systematically organized into folders based on letter classification, subsequently used to train the K-Nearest Neighbors model, known for its simplicity and effectiveness in pattern recognition tasks. The tool enables users to input Sinhala letters and immediately receive feedback on the correctness of their entries. The system evaluates the shape of the letters and provides feedback with an impressive accuracy rate of 87%, ensuring that preschoolers receive reliable insights into their learning progress. While previous studies have primarily focused on recognizing Sinhala characters using Optical Character Recognition integrated with machine learning algorithms, the proposed approach diverges by capturing letter inputs through a graphical user interface developed using Tkinter. This design significantly enhances the interactivity and user-friendliness of the tool, making it more suitable and appealing for preschool-aged children.