SUBJECTIVE ANSWERS EVALUATION USING ML & NLP

SUBJECTIVE ANSWERS EVALUATION USING ML & NLP

 Prof. Satish C. Cholke1, Shrutika Holkar2

*1 Asst. Professor Department of Information technology, Sir Visvesvaraya Institute of Technology, Nashik, Maharashtra, India.

*2 student Department of Information technology, Sir Visvesvaraya Institute of Technology, Nashik, Maharashtra, India.

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract:- Every year, educational boards and universities conduct exams in offline mode, where a large number of students take subjective type exams. Evaluating such a vast number of papers manually requires significant effort. The quality of evaluation can also vary depending on the evaluator's mood, making the process inconsistent. Moreover, the manual evaluation is both lengthy and time-consuming.

            On the other hand, competitive and entrance exams typically consist of objective or multiple-choice questions (MCQs). These exams are evaluated using machines, making the process straightforward. This method also conserves resources and minimizes human interaction, resulting in error-free evaluations. While multiple systems are available for evaluating objective (MCQ) type questions, there is a lack of provision for assessing subjective (descriptive) type questions.

            Automating the evaluation process for descriptive answers would be extremely beneficial for educational institutions, as it would enable efficient and accurate assessment of students' exam answer sheets.

Key Words:  Subjective type exams, manual evaluation, evaluation consistency, time-consuming, objective type questions, multiple-choice questions (MCQs), automated evaluation, educational institutions, efficient assessment, accurate assessment, error-free evaluations, evaluation systems, descriptive answers, resource conservation, human interaction minimization