Data provenance and trustworthiness assessment in photo archives
This event took place on Wednesday 31 May 2017 at 11:30
Art Historical Photo Archives are making their first steps in the Semantic Web. An ontology capable to represent the domain is still under development. Merging information from different data sources will highlight contradictory pieces of information – e.g. different authorship attributions of the same artwork. Such a knowledge base will give the opportunity to historians to compare contrasting data. One of the main research questions is thus how to evaluate trustworthiness by means of provenance of information, so as to enable data publishers to assess quality of their content metadata, and to provide final users of expressive ways to query such data.
In this presentation, I will first introduce my contribution to the development of ontologies for representing the Photography and Arts domains, with a focus on contradictory information related issues. Secondly, I will introduce quantitative and qualitative approaches to evaluate trustworthiness of data provided by archives. Finally, I will highlight some results and challenges in the development of a tool aimed to assess the quality of aforementioned content metadata.
Watch the webcast replay >>