Classical databases and knowledge organization: A case for Boolean retrieval and human decision-making during searches

27 Citations (Scopus)
472 Downloads (Pure)

Abstract

This paper considers classical bibliographic databases based on the Boolean retrieval model (such as MEDLINE and PsycInfo). This model is challenged by modern search engines and information retrieval (IR) researchers, who often consider Boolean retrieval a less efficient approach. The paper examines this claim and argues for the continued value of Boolean systems, and suggests two further considerations: (a) the important role of human expertise in searching (expert searchers and "information literate" users) and (b) the role of library and information science and knowledge organization (KO) in the design and use of classical databases. An underlying issue is the kind of retrieval system for which one should aim. Warner's (2010) differentiation between the computer science traditions and an older library-oriented tradition seems important; the former aim to transform queries automatically into (ranked) sets of relevant documents, whereas the latter aims to increase the "selection power" of users. The Boolean retrieval model is valuable in providing users with the power to make informed searches and have full control over what is found and what is not. These issues may have significant implications for the maintenance of information science and KO as research fields as well as for the information profession as a profession in its own right.

Original languageEnglish
JournalJournal of the Association for Information Science and Technology
Volume66
Issue number8
Pages (from-to)1559-1575
Number of pages17
DOIs
Publication statusPublished - 1 Aug 2015

Keywords

  • Faculty of Humanities
  • klassiske databaser
  • Boolesk søgning
  • søgemaskiner

Fingerprint

Dive into the research topics of 'Classical databases and knowledge organization: A case for Boolean retrieval and human decision-making during searches'. Together they form a unique fingerprint.

Cite this