INTELLIGENT SYSTEM FOR DESCRIPTIVE ANSWER EVALUATION

Abstract

Exams are a reliable and well-liked method of assessing the amount of information students have learnt. Exam evaluation is a difficult and time-consuming undertaking. Teachers will find this

assignment handy since it requires less time and effort thanks to the automated response script as- sessment. Despite the fact that instructors may now give computerised objective and short answer

tests, it may be difficult to evaluate descriptive responses. The computer must take into account the crucial elements that a human assessor takes into account while manually examining descriptive replies. This research attempt offered an intelligent assessment platform that takes into account the question type, relevant keywords, structural, conceptual, and linguistic factors when analysing

a response in order to get around these difficulties. Concept graphs, fuzzy string matching, gram- mar checking, and other natural language processing similarity metrics were used to assess the

proposed model. For instructors and students, the system also creates customised feedback and analysis reports to support focused learning.