Skip to content

Semantor

Semantor is a research project that tries to make use of the growing number of Linked Data sources in order fo find experts. In our main scenario, the user starts with a topic for expert search and wants to explore the Linked Data traces that users have left on the Web, and that have something to do with this topic.

In order to ensure the best results of expert search Semantor research is set to answer the following research questions:
- how to choose the expertise hypothesis for expert search, based on the structure od the dataset(e.g., use the expertise hypothesis ” people who wrote at least two tweets on the topic are experts in the topic” for one data set and ” people who wrote a scientific publication on the topic are experts in the topic” for another dataset)?
- how to explore similar topics and get to experts in adjacent feels using the semantic background knowledge bases such as DBPedia?
- how to exploit granularity of topics used with user traces in order to rank the expertise (experts are supposed to use more specific terms then non-experts)?

The preliminary results of this research project can be found in the following publications:

  • Stankovic, M., Open Innovation and Semantic Web : Problem Solver Search on Linked Data. In Proceedings of International Semantic Web Conference (ISWC) 2010; 7th-11th Novebmer, 2010, Shanghai, China
  • Stankovic, M., Wagner, C., Jovanovic, J. and Laublet, P., Looking For Experts? What can Linked Data do for You? In Pre-proceedings of Linked Data on the Web 2010 (LDOW) workshop, within WWW2010 conference; 26th-30th April, 2010, Raleigh, NC, USA

and in the following Web articles:

Semantor is being developed in cooperation with the STIH laboratory from the University Paris-Sorbonne.