OMNI--Alternative Approaches to Internet Metadata.

The growth in the size of the Internet has resulted in much effort being spent on indexing its contents. The most popular solutions are created by automatic methods, and although offering impressive coverage, they are disappointing where precision of meaning is required. Alternative services created by human beings arrange and index resources according to concept and offer more relevant retrieval, but cannot hope to achieve 100% coverage. New developments are imminent which may support both the automated and non-automated approaches, and vastly improve the quality of Internet metadata. This paper demonstrates that the task of locating information on the Internet has not been and will not be accomplished by the use of sheer computing power alone. It reviews the topic of metadata creation, with special reference to the OMNI project (Organizing Medical Networked Information) and describes two encouraging new initiatives: PICS (Platform for Internet Content Selection) and the Dublin Core Metadata Set/Warwick Framework. All users need improved search tools with which to navigate the Internet. The answers to the present dilemma will be based on more than one tool, on both human intervention and intelligent automated data gathering. Metadata, as exemplified by the old catalog card, is set to become a key concept in the future of the new information world. (Contains 19 references.) (Author/SWC)