International Journal of Advances in Computer Science and Its Applications
Author(s) : ANBUSELVAN SANGODIAH, LIM EAN HENG, ROHIZA BT AHMAD, WAN FATIMAH BT WAN AHMAD
At present, with the growing demand of knowledge economy, fast popularization and development of the World Wide Web, a lot of people are turning to e-learning via web particularly in educational institutions. There are plenty of elearning systems nowadays and these systems provide functionalities ranging from some simple tools to manage and search a lot of teaching materials to knowledge sharing through online forum discussion. While the online forum discussion promotes exchanging ideas between instructors and students, however due to the job demand of instructors at their workplace and at times being away from their workplace would often result in responses to the students’ questions were not being responded in time. To address the problem, automated response systems in forum discussion have been developed where the system would take the responsibility to respond to questions posed by student in forum discussion in the absence of instructors thus reducing the workload of instructors significantly. Despite some research work has been revolved around automated response system in forum discussion in e-learning system, the extent to which the depth of responses or knowledge in accordance to Blooms taxonomy has not been fully explored. Equally important, besides matching keywords of questions to generate responses is the ability to provide adequate details of knowledge to a particular question in the order of depth of knowledge in the context of Blooms taxonomy. This paper will highlight an integrated approach to extract relevant knowledge in accordance to Blooms taxonomy from learning materials particularly in forum discussions (software forum threads) in response to questions posed by learners. As a result, an enhanced architecture of question and answering system in the context of forum discussion would be proposed. Also will be discussed are the technologies such as text mining, data mining and artificial intelligence to accomplish the proposed tasks. On completing the suggested tasks, a more intelligent automated response system in accordance to Bloom taxonomy can be produced.