Towards a Similarity-Based Identity Assumption Service for Historical Places

Acquisition and semantic annotation of data are fundamental tasks within the domain of cultural heritage. With the increasing amount of available data and ad hoc cross linking between their providers and users (e.g. through web services), data integration and knowledge refinement becomes even more important. To integrate information from several sources it has to be guaranteed that objects of discourse (which may be artifacts, events, persons, places or periods) refer to the same real world phenomena within all involved data sources. Local (database) identifiers however only disambiguate internal data, but fail in establishing connections to/between external data, while global identifiers can only partially solve this problem. Software assistants should support users in establishing such connections by delivering identity assumptions, i.e. by estimating whether examined data actually concerns the same real word phenomenon. This paper points out how similarity measures can act as groundwork for such assistants by introducing a similarity-based identity assumption assistant for historical places to support scholars in establishing links between distributed historical knowledge.