Design and Development of an AI-Assisted Web Learning Platform for Exam Planning, Self-Assessment and Personalized Gap Detection
TL;DR
The experimental evaluation shows an improvement in student performance between the first and last attempt, high structural validity of the generated questions, concise summaries and low response times for data access operations, which support the usefulness of the platform as a complementary tool for active learning, self-assessment and personalized identification of knowledge gaps.
Abstract
In the context of the increasing volume of digital learning materials used in education, students often face difficulties in organizing content, identifying knowledge gaps and maintaining an active learning process. This paper presents SmartStudy AI, a web-based educational platform that transforms course materials into AI-assisted learning resources. The system supports document upload, summary generation, multiple-choice question generation, flashcards, formative assessment and personalized recommendations. The platform is implemented using Java EE, JSP, Jakarta Servlets, MySQL and direct integration with the Gemini 2.5 Flash API. The experimental evaluation, conducted on a reproducible corpus of users, courses, questions and quiz attempts, shows an improvement in student performance between the first and last attempt, high structural validity of the generated questions, concise summaries and low response times for data access operations. The results support the usefulness of the platform as a complementary tool for active learning, self-assessment and personalized identification of knowledge gaps.