Quertle

Quertle is a semantic search engine for life and chemical science literature and information.^[1][2][3][4] It covers a wide variety of information sources.

How Quertle Works

Quertle uses semantic-based linguistics to automatically extract subject–verb–object relationships asserted by the author(s) of each document. The identification of these assertions uses several methods including natural language processing.^[5][6] For full-text documents, Quertle includes only the main content, not, for example, the references.

The subject–verb–object relationships are stored in a metadatabase and the user's query is matched against that metadata. This identifies documents based on meaning and context and generally provides fewer, but more relevant, hits than a traditional keyword search. Thus, Quertle is fundamentally different from search sites such as PubMed. Nonetheless, Quertle does simultaneously search a keyword index to find documents based on inclusion of the search terms. These are presented on a separate tab in the results.

An ontology covering genes, proteins, chemicals, diseases, cell types, and other life, chemical, and biomedical science nomenclature is used to automatically search for all variants of a term in the user's query. For example, a search for "aspirin" will find asserted relationships that mention "acetylsalicylic acid". The ontology also is used to find members of a class of entities, such as "neurotransmitters".^[7]

Content

Quertle indexes MEDLINE, full-text articles from BioMed Central^[8] and PubMed Central (open access subset), NIH grants, the US National Library of Medicine TOXNET database, and biomedical news.^[9]

Criticism

It has been suggested that details of how Quertle works ‘are not clearly described to the public’.^[7]

References