Application of immune network metaphor to keyword map-based topic stream visualization

A method to find topic distribution from a sequence of document sets is proposed. As the Web becomes one of the most important information resources for us, the interaction between a human and a Web interface should be considered from various viewpoints, besides that of document retrieval. In this paper, we focus on the extraction and visualization of topic distribution over the Web. The proposed clustering method employs the immune network model, in which the property of memory cell is used to find the topical relation among document sets. The effectiveness of the proposed method is also shown by applying it to two sequences of online news articles. Furthermore, keyword map-based information visualization system is developed to visualize the topic streams found from a sequence of document sets.