Usage and Impact of Controlled Vocabularies in a Subject Repository for Indexing and Retrieval

Abstract

Since 2009, the German National Library for Economics (ZBW) supports both indexing and retrieval of Open Access scientific publications like working papers, postprint articles and conference papers by means of a terminology web service. This web service is based on concepts organized as a ‘Standard Thesaurus for Economics’ (STW), which is modelled and regularly published as Linked Open Data. Moreover, it is integrated into the institution’s subject repository for automatically suggesting appropriate key words while indexing and retrieving documents, and for automatically expanding search queries on demand to gain better search results. While this approach looks promising to augment ‘off the shelf’ repository software systems in a lightweight manner with a disciplinary profile, there is still significant uncertainty about the effective usage and impact of controlled terms in the realm of these systems. To cope with this, we analyze the repository’s logfiles to get evidence of search behaviour which is potentially influenced by auto suggestion and expansion of scientific terms derived from a discipline’s literature.