View article

Hypertext information retrieval systems combine hypertext and information retrieval capabilities by providing retrieval techniques that include direct search as well as navigation and browsing. The incorporation of a mechanism for the automatic construction of hypertext into an IR system is quite important in order to exploit the advantages of the integration of hypertext with information retrieval.

News hypertext deals with articles from newspaper archives. Formal ways to capture the temporal aspects that characterize the newspaper domain, and which are often ignored by IR systems, are presented. The suggested framework is conceptualized with the notions of stories and threads, which are substories within a story. Threads are identified by applying clustering techniques to articles’ segments that correspond to subtopics within the main topic of an article and then linking the segments which belong to the same cluster.