Multi swarm optimization based automatic ontology for e-assessment

Abstract

The utilization of ontology in the e-assessment area has grown tremendously. The context of e-learning is significant to the students for educational purposes. This makes the testing process easy for the students and also for the teachers. The majority of the approaches that deals with the ontology issue have suggested that the individual ontology models have merely a fraction of the assessment domain. To trounce such drawbacks, here, an automated ontology creation is proposed for the e-assessment systems. Initially, the text is extracted from the web utilizing the Unsupervised Quick Reduct (UQR) algorithm. This is trailed by the summarization of the texts using the multi-swarm optimization (MSO) based on preference learning. Finally, the sentence of the summary is then transmuted to multiple choice questions (MCQ). The keys are created using statistical pattern (SP). The efficiency of the system is examined using the experimental outcomes like error rate, precision, recall and accuracy. In accuracy, the proposed UQR algorithm achieves 97.7%, MSO achieve 96.2% accuracy and key generation achieves 94.7% accuracy. The proposed automatic ontology system indicates better when weighed against the top-notch methods.