Towards quality measures for evaluating thesauri

Abstract

Thesaurus evaluation has often focused on the task of retrieval or analyzed the information environment including factors such as the user, the user interface. Rarely ever has the quality of a thesaurus as an artefact itself been focus of attention. In particular, there are no measures, which provide a holistic description of its intrinsic characteristics as an artefact. This paper suggests a range of abstract measurement constructs based on quality notions in thesaurus literature and inference from other related literature. The suggested measurement constructs allow representing a thesaurus in evaluation approaches, which thereby becomes empirically testable. © 2010 Springer-Verlag Berlin Heidelberg.