Executive Summary : | The University Grants Commission (UGC) and the All India Council for Technical Education (AICTE) have proposed Outcome-Based Education (OBE) for all higher education disciplines to reform the Indian education system. This involves identifying learning outcomes in terms of Course Outcome (CO), Program Specific Outcome (PSO), and Program Outcome (PO) and mapping CO to PO and PSO. To achieve these outcomes, test questions should be mapped to CO and direct mapped to different cognitive levels of Bloom's taxonomy. However, this task requires training, skills, and expertise. To support OBE in a standardized manner, tools based on Natural Language Processing (NLP) and Machine Learning (ML) are particularly beneficial. NLP can model linguistic behavior and semantic relationships among text elements, while ML is used for predictive analysis, regression, classification tasks, topic modeling, and text summarization. In education technology, human analysis of text complexity is a time-consuming and challenging process. Automated text evaluation and scoring systems for descriptive or essay-type answers provide real-time feedback for students to improve their answers. However, most contemporary tools offer only linguistic support and do not consider the subject context. An NLP and ML-based approach could be a solution to these problems.
The research proposes designing an educational support system using NLP and ML to guide teachers in setting questions at different cognitive levels of Bloom's taxonomy, mapping questions to their corresponding CO, finding appropriate topics for question setting, and identifying text complexity in learning materials development. The system will also help students provide formative feedback on descriptive answers by auto-grading response text. |