Support : Publications : Chronolog Archives : September/October 2004

From the Experts...
Thesaurus Mapping Feature on Dialog DataStar

Are you unfamiliar with the terminology or descriptor terms from MeSH® (MEDLINE® thesaurus for MEDL and AMED), EMTREE (EMBASE® thesaurus for EMED) or the other vocabulary files associated with BIOSIS (BIOL), Cinahl® (NAHL), PsycINFO® (PSYC), Allied & Complementary Medicine™ (AMED), and British Nursing Index (BNID)? Do you need to verify a descriptor? Have you not received the desired result in your free-text search? If so, the Thesaurus Mapping tool can help you because it links a free-text term to the most relevant controlled vocabulary descriptor from the database.

The thesaurus descriptors are controlled terms, which have been hierarchically structured and are used for subject indexing. The Thesaurus Mapping tool analyzes the documents searched and ranks the most relevant descriptors. The regular thesaurus in Dialog DataStar databases only lists the descriptors. For example, if you search for MMR Vaccine in free text in MEDLINE and map the terms, Dialog DataStar will retrieve "Measles, Mumps and Rubella Vaccine" from the thesaurus.

View Figure 1

The Thesaurus mapping feature is particularly useful when terms entered directly into the database do not retrieve relevant hits: for example, in MEDLINE, skin cancer is not considered a MeSH term and retrieves very few results. However, if you check the Thesaurus mapping box on the DataStarWeb search screen and enter the term "skin cancer," Dialog DataStar will retrieve "Skin-Neoplasms" at the top of the list of mapped terms. If you search on this term, the number of results will be considerably greater and more precise because Dialog DataStar has mapped the descriptors in the documents and presented them in order of relevance. Once you have obtained your preferred term, you can use the same features that are available in the Thesaurus: explode, major and subheadings. See Figure 1.

Note: Ranking should not be confused with Thesaurus Mapping. Ranking analyzes statistical trends found in your search results by counting the occurrences of unique terms within a specific field from your search result. In some databases, you can rank on descriptor terms, but this will only give you the number of occurrences in which this term has appeared in the number of documents selected.

If you rank on the first 100 descriptors in databases that offer a Thesaurus only, you will receive the same ranked list of Thesaurus terms as if you had used the Thesaurus Mapping tool for descriptor terms only.

Leirae Ramirez
Bern, Switzerland

top