Any designer of a digital library must consider short-term practical access issues like how the library fits into current de-facto standards for internet access (such as email, ftp, archie, gopher, etc), and long-term access issues like how the huge volume and rapid change of the DL information can be made manageable to potential users. There is no doubt that, long term, the lack of intelligent and intuitive tools for DL access will turn people away from using them.
This paper discusses, in brief, the issue of intelligent access to on-line information, such as digital libraries, in the context of two ongoing efforts: The ACM SIGART Electronic Information Service, and the ACM Global Electronic Community. The paper begins by discussing these two efforts and the perceived need for intelligent information retrieval. It then presents some arguments for using a knowledge representation approach to information retrieval over a database approach, and concludes with a sample of the ontology being developed for the SIGART EIS.