COGNILY: A CONTEXT-AWARE, BLOOM’S TAXONOMY-ALIGNED QUESTION PAPER GENERATION SYSTEM FOR ENGINEERING UNIVERSITIES

Abstract

The growing pressure on outcome-based learning in engineering colleges has highlighted the necessity to have smart systems that will produce quality, balanced and pedagogically sound question papers. This paper introduces Cognily, a history-conscious and Taxonomyaligned Bloom question generation system that aims to streamline and optimize the assessment creation process. The suggested system incorporates past question paper information, course marking, and categorization of cognitive level to create and produce a wide variety of different and non-repetitive examination papers. Cognily uses machine learning and rule-based to process previously asked questions, find patterns, and avoid redundancy without compromising on the right level of difficulty. The system divides questions into Bloom levels of Taxonomy- Remember, Understand, Apply, Analyze, Evaluate and Create- to cover all the areas of assessment. It also allows the faculty to create question papers based on academic needs, as it supports customizable constraints like subject weightage, unit-wise distribution and difficulty balance. The implementation proves to be more efficient, less manual and more fair in terms of assessment design. Cognily helps to make the evaluation process in engineering education more effective and standardized by making sure that it is aligned with the learning objectives and by reducing repetition.