CogniTwin: AStress-Adaptive Summarizer and Dynamic Quiz Generator for Programming Education

Author:

Pasupula Narasimha

^{Central University of Andhra Pradesh, Ananthapuramu}

Abstract - Nowadays, students and professionals are bombarded by an overwhelming volume of textual information in the digital world, leading to cognitive overload and reduced understanding. Static outputs from traditional summarization systems, and generic question banks from quiz generators lack personalization and relevance respectively. CogniTwin solves these problems by incorporating a stress-adaptive summarizer and a dynamic quiz generator into one platform. Using transformer based models, it generates summaries that adapt to learner’s stress levels to reduce cognitive load. The use of subject-relation-object triplet extraction creates a structured knowledge graph that helps produce exams with relevant distractors and provides explanatory feedback. The export function supports professional documentation with PDF and DOCX. The results showed higher engagement and comprehension than existing static tools, and CogniTwin aims to bridge theory and practice of AI assisted learning technology.

Key Words: Cognitive Load, Self-Efficacy, Summarization, Quiz Generation, AI-Assisted Learning, Educational Technology.