Algorithms for extracting remarkable points on the Web focusing on topic transition processes aimed at efficient explore support

In this research, a method to effectively support Web exploring is developed. A promising approach is extraction of topic transition processes on the Web and their visualization. Specifically, this research develops algorithms to cooperatively extract 1) Remarkable points in topic transitions among Web documents and 2) Border lines for clustering Web documents in accordance with topics. This paper mainly describes an outline of the proposed method and an algorithm for extracting the topic changing points. This paper also describes an experiment to evaluate the algorithm. Finally, feature and basic effectiveness of our method are discussed based on the results of the experiment.