Web Text Clustering for Personalized E-learning Based on Maximal Frequent Itemsets

With the rapid development of the network technique and the prevalence of the Internet, e-learning has become the major trend of the development of international education since 1980s, and the important access for the internationalization and the information of education. To meet the personalized needs of learners in e-learning, a new Web text clustering method for personalized e-learning based on maximal frequent itemsets is proposed. Firstly, the Web documents are represented by vector space model. Then, maximal frequent word sets are discovered. Finally, based on a new similarity measure of itemsets, maximal itemsets are used for clustering documents. Experimental results show that the proposed method is effective.