A Framework for Semantic Driven E-examination System for Subjective Questions,

Abstract

In this paper, a theoretical framework that incorporates a semantic dictionary (WordNet) into the current e-examination system to support subjective question is proposed. The WordNet serves as a medium of enhancing the current e-examination system in which machines can process, interpret, and evaluate information for enhance knowledge retrieval. The approach adopted is hypothesized to solve the semantic mismatch issues in subjective e-examination. An algorithm that explores the synset structure of WordNet database and which gives meaning of any term based on the available synonyms is proposed. The framework is implemented using Java programming language and a prototype of the proposed system is tested and compared with the existing system. Results show that words that are synonymous to any given correct answer are equally recognize as correct option. Hence, the e-examination system reliability, performance and assessment of depth of knowledge of students have been greatly increased.