Cultural organizations publish independent from each other information about their collections on the internet. For example, museums provide information about impressionist paintings, while other organizations provide information about impressionist painters. With current research methods, this information is only available in isolated form and there is no cohesion. As a result, it requires a lot of manual search work. In his thesis, ‘End-user support for access to heterogeneous linked data’, Michiel Hildebrand from the Centrum Wiskunde & Informatica (CWI) in Amsterdam, investigates how information from different sources in the cultural heritage domain can be made accessible for the end-user.
The development of the Semantic Web enables data from different sources to be related to each other. The research of Hildebrand is part of MultiMediaN e-culture, a national e-culture project. In this project, collections of Dutch museums, i.e. the Rijksmuseum and extensive background information is used from the cultural heritage domain. On April 20, Michiel Hildebrand will defend his thesis at the University of Amsterdam. Results of his research can be applied by experts in the cultural heritage domain, music industry and news agencies.
A big challenge in using these data sources is their heterogeneity. Not only is there a large diversity in data sources, but also the way in which they are described has large variations. Much information is available about the Impressionist form of art, such as related painters, paintings, specific materials, techniques, stories and places. In addition, the organizations all pay different attention to, for example, a famous impressionist painter like Monet.
From the perspective of the end-user, Hildebrand developed different search algorithms and presentation methods that could handle heterogeneous linked data. In his research there are three central questions: What data is available and what are their relations to one other? Which algorithms are flexible enough to search in heterogeneous sources? Which presentation methods are suited to support end-users in the search process? It is the first time that research in this field includes all these aspects.
Research from Hildebrand shows that it is possible to develop the support in the search process by combining information from different data sources by using technologies for the Semantic Web. By working with experts from the cultural heritage domain, it becomes clear that these technologies can be used to better support users with their daily search tasks in large amounts of linked data.
Various new questions arise with the explosive amount of digital data offered to society and science. This is an important research theme at CWI. The research from Hildebrand has provided a contribution to this area.