Change Summarization in Web Collections

World Wide Web is not only enormous but also dynamic information space. Every day large quantity of new information is published on web pages. Many times people want to know what are the major changes in their area of interest over a given time period. This paper addresses the problem of summarizing changes in web collections devoted to a common topic. We have created a system called ChangeSummarizer, which periodically monitors a web collection in search for new changes and generates their summary. Since many web pages can be quite static over long time or just unrelated to the query we employ the method to evaluate, which pages are dynamic and which provide valuable content. Basing on this evaluation ChangeSummarizer creates web page ranking list and updates it regularly in order to improve subsequent summaries. Additionally, the system searches for new, valuable web pages, which can be included into the collection to enhance its quality.

• Type: Conference
• Year: 2004
• Where: IEAAIE
• Authors: Adam Jatowt, Khoo Khyou Bun, Mitsuru Ishizuka
• Keywords: IEAAIE 2004, Web Collection, Web Pages, World Wide Web