Optimal document-indexing vocabulary for MEDLINE

A detailed analysis of prior results and their underlying experimental designs indicates that there are a number of open questions relevant to the overall debate on indexing vocabularies for MEDLINE and results from a new experiment assessing eight different retrieval strategies indicate that MeSH does have an important role in text retrieval.

Abstract

The indexing vocabulary is an important determinant of success in text retrieval. Researchers have compared the effectiveness of indexing using free-text and controlled vocabularies in a variety of text contexts. A number of studies have investigated the relative merits of free-text, MeSH and UMLS Metathesaurus indexing vocabularies for MEDLINE document indexing. Most of these studies conclude that controlled vocabularies offer no advantages in retrieval performance over free-text. This paper offers a detailed analysis of prior results and their underlying experimental designs. The analysis indicates that there are a number of open questions relevant to the overall debate on indexing vocabularies for MEDLINE. This paper also offers results from a new experiment assessing eight different retrieval strategies. These strategies involve document indexing via free-text, MeSH and several alternative combinations of the two vocabularies. The results indicate that MeSH does have an important role in text retrieval.