J. Beach, S.N. Minton, and W.E. Rzepka (USA)
Software Agents, Learning Algorithms, High Speed Internet, Information Infrastructures, Internet Tools, Internet Access.
This paper describes a Web data extraction system and how it was used to satisfy several, distinct information needs. The Web extraction capability is based on software agent and learning algorithms which are trained to navigate through Web sites and extract specific data from Web pages. An infrastructure for scheduled execution of the software agents is described. The application of agent scheduling technology for the development of timely reports which aggregate and fuse information from several disparate Web data sources is described. These reports are provided on a periodic basis and can be used to monitor web sites for specific changes in their content. The changes are reported to users via email.
Important Links:
Go Back
