Topic Segmentation of E-learning materials using Domain and Pedagogical Ontology

摘要：

Coherent segmentation of e-learning materials according to topic is an important task in e-learning systems for information extraction. This paper proposes a generative statistical model to segment elearning materials from a topic modeling perspective. Latent Dirichlet Allocation (LDA) is a powerful statistical method to extract topics from the documents. It has been applied to many text and multimedia applications for information extraction. This paper proposes the usage of hierarchical LDA model with both domain ontology based on ACM classification and pedagogical ontology for text segmentation of e-learning materials. The proposed model assigns each segment to a particular topic based on domain and pedagogical ontology for categorizing the corresponding document to potentially multiple topics.

关键词： component Latent Dirichlet Allocation(LDA) Ontology Pedagogy

作者: K.Sathiyamurthy T.V.Geetha

作者单位: Department of Computer Science & Engineering Anna University Chennai India Department of Computer Science and Engineering Anna University Chennai India

会议类型: 国际会议

会议名称: 2010 International Conference on Future Information Technology(2010年未来信息技术国际会议 ICFIT 2010)

会议地点: 长沙

会议语种:英文

页码: 52-56

在线出版日期: 2010-12-14（万方平台首次上网日期，不代表论文的发表时间）