KINDEX – Automatic Subject Indexing with Knowledge Graph Services

Automatic subject indexing has been a longstanding goal of digital curators to facilitate effective retrieval access to large collections of both online and offline information resources. Controlled vocabularies are often used for this purpose, as they standardise annotation practices and help users to navigate online resources through following interlinked topical concepts. However, to this date, the assignment of suitable text annotations from a controlled vocabulary is still largely done manually, or at most (semi-)automatically, even though effective machine learning tools are already in place. This is because existing procedures require a sufficient amount of training data and they have to be adapted to each vocabulary, language and application domain anew. Against the background of tight budgets and missing IT personnel in cultural heritage as well as research infrastructure institutions, adoption of automatic subject annotation tools is hindered, while manual assignment of index terms is an even greater burden on the available financial resources. In this paper, we argue that there is a third solution to subject indexing which harnesses cross-domain knowledge graphs (i.e., DBpedia and Wikidata) to facilitate cost-effective automatic descriptor assignments that can be done without any algorithm tuning and training corpora. Our KINDEX approach fuses distributed knowledge graph information from different sources. Experimental evaluation shows that the approach achieves good accuracy scores by exploiting correspondence links of publicly available knowledge graphs.