manuscript pdf

manuscript

More Book Details

Description of the Book:

Ontologies ncontain knowledge in the form of concepts, predicates, instances and ntheir relationships. This knowledge can be exploited by an assessment nsystem in the form of multiple choice questions (MCQs). Most of the nexisting MCQ generation approaches are limited to questions of type Whatn is C ? or Which of he following is an example of C ? (where C is a nconcept symbol). Besides the question types, the methods for generating ndistracting answers (distractors) are to be given much importance. Thesen distractors often determine the quality and difficulty level of a nmultiple choice question. Currently, there are no systematic methods forn generating distractors of a MCQ from an ontology. In this paper, we npropose two MCQ generation approaches and the corresponding distractor ngenerating techniques. Our distractor generation techniques, unlike nother methods, consider Open World Assumption, so that the generated nMCQs will be always precise (ensures falsity of distracting answers).

Inn addition, we present a metric to calculate the difficulty level (a nvalue between 0 and 1) of the generated MCQs. The proposed system is nimplemented, and experiments on pecific ontologies have shown the neffectiveness of the approaches. Our results show that the proposed napproaches, when used with semantically-rich ontologies, can provide a nset of quality MCQs on the domain.’, ‘Ontologies ncontain knowledge in the form of concepts, predicates, instances and ntheir relationships. This knowledge can be exploited by an assessment nsystem in the form of multiple choice questions (MCQs). Most of the nexisting MCQ generation approaches are limited to questions of type Whatn is C ? or Which of he following is an example of C ? (where C is a nconcept symbol). Besides the question types, the methods for generating ndistracting answers (distractors) are to be given much importance. Thesen distractors often determine the quality and difficulty level of a nmultiple choice question.

Currently, there are no systematic methods forn generating distractors of a MCQ from an ontology. In this paper, we npropose two MCQ generation approaches and the corresponding distractor ngenerating techniques. Our distractor generation techniques, unlike nother methods, consider Open World Assumption, so that the generated nMCQs will be always precise (ensures falsity of distracting answers). Inn addition, we present a metric to calculate the difficulty level (a nvalue between 0 and 1) of the generated MCQs. The proposed system is nimplemented, and experiments on pecific ontologies have shown the neffectiveness of the approaches. Our results show that the proposed napproaches, when used with semantically-rich ontologies, can provide a nset of quality MCQs on the domain

• Creator/s: Vinu E.V
• Date: 5/5/2014
• Book Topics/Themes: Ontologies, semantic web, multiple choice questions, automatic question generation

An excerpt captured from the PDF book