The Indexing Initiative A Report to the Board of Scientific Counselors of the Lister Hill National Center for Biomedical Communications

For more than 150 years, the National Library of Medicine has provided access to the biomedical journal literature through the analytical efforts of human indexers. Since 1966, access has been provided in the form of electronically searchable document surrogates consisting of bibliographic citations, descriptors assigned by indexers from the MeSH ® controlled vocabulary (MeSH, 1998) and, since 1974, author abstracts of many, but not all, items. In the late 1990s, as medical journals migrate from print to electronic form, the need for human intervention to link users with relevant documents may be minimized, if not eliminated altogether. In addition, the cost of human indexing of the biomedical literature is high. As budgets are reduced and costs continue to climb, it seems reasonable to investigate alternative methods for indexing bibliographic and other data. The MEDLINE ® database contains about 11 million records, all of which have been produced by human indexing. The file presently grows at the rate of about 400,000 indexed citations per year, covering about 4,300 international biomedical journals. Human indexing consists of reviewing the complete text of each article, rather than an abstract or summary of it, and assigning descriptors that represent the central concepts as well as every other topic that is discussed to a significant extent. Indexers assign descriptors from the MeSH vocabulary of more than 19,000 main headings. Main heading descriptors may be further qualified by selections from a collection of 88 topical subheadings. Since 1990, there has been a steady and sizeable increase in the number of articles received, owing both to an increase in the number of indexed journals and, to a lesser extent, to an increase in the number of articles in journals that are already being indexed. In the face of a growing workload and dwindling resources, we have undertaken the Indexing Initiative to reexamine both the way that MEDLINE is currently produced and also the ways in which NLM might accomplish its mission of providing access to biomedical literature other than by manual subject indexing. The objective of NLM's Indexing Initiative (IND) is to investigate methods whereby automated indexing methods partially or completely substitute for current indexing practices. The project will be considered a success if methods can be designed and implemented that result in retrieval performance that is equal to or better than the retrieval performance of systems based principally on humanly assigned index terms. Human indexing is an …