Bibliographic data mining: Automatically building component part records for E-Journal articles on the Internet
Ying Li, Dick R. Miller, Mary Buttner · 2002
This paper is based on a presentation at the Medical Library Association Annual Meeting in May 2000. With the rapid expansion of fulltext on the World Wide Web, Internet navigation has become a serious concern to libraries. To increase access to fulltext digital resources and minimize staff effort, Stanford's Lane Medical Library began a Bibliographic Data Mining project in February 1999. With limited human editing, the PERL scripts captured relevant URLs from selected e-journal Web sites and automatically created basic MARC component part bibliographic records for the articles in those journals. This project demonstrated that bibliographic data mining is a feasible methodology for improving the effectiveness of access to Internet resources within the context of the online catalog. It provided valuable experience in managing digital resources and highlighted the need for more consistent metadata.