Scientometrics

, Volume 102, Issue 3, pp 2323–2345

Science models for search: a study on combining scholarly information retrieval and scientometrics

Article

DOI: 10.1007/s11192-014-1485-2

Cite this article as:
Mutschke, P. & Mayr, P. Scientometrics (2015) 102: 2323. doi:10.1007/s11192-014-1485-2

Abstract

Models of science address statistical properties and mechanisms of science. From the perspective of scholarly information retrieval (IR) science models may provide some potential to improve retrieval quality when operationalized as specific search strategies or used for rankings. From the science modeling perspective, on the other hand, scholarly IR can play the role of a validation model of science models. The paper studies the applicability and usefulness of two particular science models for re-ranking search results (Bradfordizing and author centrality). The paper provides a preliminary evaluation study that demonstrates the benefits of using science model driven ranking techniques, but also how different the quality of search results can be if different conceptualizations of science are used for ranking.

Keywords

Information retrieval Scientometrics Bradfordizing Network analysis Betweenness Science models 

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2014

Authors and Affiliations

  1. 1.GESIS–Leibniz-Institute for the Social SciencesCologneGermany

Personalised recommendations