In the abstract to this article the authors state, "This suggests that full-text searching alone may be sufficient, and that metadata searching as a surrogate is not necessary."
However, in their conclusion, they admit, "One limitation...
Read more...
In the abstract to this article the authors state, "This suggests that full-text searching alone may be sufficient, and that metadata searching as a surrogate is not necessary."
However, in their conclusion, they admit, "One limitation of the study is that searching by gene name may not be representative of general biomedical literature searches. Gene names are specific terms for concepts and, as a result, can be searched for more easily in full-text compared to other concepts in the biomedical literature. Concepts that are referred to by many different terms are why controlled vocabularies such as MeSH (2007) are so helpful in the biomedical domain. In addition, gene names may occur more frequently in the full-text compared to the metadata, as compared to other types of terms. Further, the searching conducted in this experiment is simplified in that only exact matching is performed using basic search functions (MySQL full-text search)."
So, the balance of the article actually contradicts the stated "conclusion" that is in the article's abstract. The authors found that in a very narrow domain, gene names, full-text searching sometimes is okay. This domain does not have much variation in terminology, so the synonym problem doesn't occur much there. So, the article's conclusions are not really meaningful outside of this narrow domain; they cannot be generalized. Thus the abstract's stated conclusion is misleading. This article adds very little to the discussion of comparing the strengths and weaknesses of metadata-enabled and full-text searching.
