INTEGRATED TASK GENERATION, EXECUTION, AND ASSESSMENT METHODS FOR ENHANCING PERSONALIZED LEARNING

The increasing use of artificial intelligence (AI) in education has enabled automated creation of learning tasks.While GenAI shows promise in generating text-based exercises, a gap remains in bridging the generated content with executable environments.Existing approaches primarily focus on task generation and often lack reliable mechanisms for execution and assessment.This paper proposes an enhanced, integrated approach of the automated task generation method, the automated deployment method and the automated assessment method by leveraging the capabilities of GenAI, which, unlike existing approaches, enables the control of task generation through new LTDL-validation mechanisms.The method combines three interrelated methods within a unified studyflow and addresses key limitations of the current AI-driven educational system.This study aims to enhance the design, execution, and assessment of practical learning tasks through a unified, AI-supported method that ensures both the semantic quality of the task and the technical validity of the execution environment.The proposed methodological framework is mathematically formalized through a mapping function that translates formal task descriptions into environment configuration instructions, a vector-based model representing the virtual environment state, and a boolean verification function utilizing cryptographic hashing for precise, automated assessment of student progress.Experimental evaluation focuses on technical feasibility and scalability, including the reliability and performance of automated environment deployment and assessment monitoring under realistic infrastructure conditions.The results demonstrate that the formalization overhead is minimal, the method is scalable, and the primary performance bottleneck lies in environment provisioning rather than in the logic of the proposed approach.