Automated Subjective Answer Evaluation System Using ML.

Evaluating subjective answers in academic assessments is often a time-consuming and inconsistent process.Traditional manual grading methods suffer from subjectivity, potential biases, and scalability issues.To address these challenges, this research presents an Automated Subjective Answer Evaluation System leveraging Machine Learning (ML) techniques.The proposed system utilizes Natural Language Processing (NLP) and semantic similarity models to assess answers based on their relevance, coherence, and alignment with the expected responses.Key components include text preprocessing, feature extraction, and similarity computation using models such as TF-IDF and transformer-based architectures.The system is trained on a diverse dataset containing expert-graded answers to improve accuracy and adaptability across various subjects.Experimental results demonstrate that the proposed approach achieves high correlation with human grading, reducing evaluation time while maintaining reliability.This research contributes to the field of automated assessment systems, enabling scalable and unbiased subjective answer evaluation in educational environments.