Special Interest Groups Discovery and Semantic Navigation Support within Online Discussion Forums

Online discussion forums provide open workspace allowing learners to share information, exchange ideas, address problems and discuss on specific themes. But the substantial impediment to its promotion as effective e-learning facility lies in the continuously increasing postings but with discrete and incoherent structure as well as the loosely-tied learners with response-freeness. This paper proposes a hybrid approach to automatically discover special interest groups within discussion forums. Once a learner becomes a member of a special interest group, he will be informed of other learning companions to enhance their in-depth communication and learning, and the newly-emerged related information will be proactively pushed to him as well. Furthermore, by identifying the posting themes and types, this paper presents a semantic search to assist learners navigating through well-structured and coherent postings to meet their learning demands. The proposed approach has been integrated into a discussion forum, and the experimental results show that the approach is feasible and efficient, enabling the effective discovering of interest groups and proper demand-driven navigational guidance.