Section: New Results

User Oriented Expert Finding

Participants : Elena Smirnova, Brigitte Trousse.

This work takes place in the context of Elena Smirnova's Ph.D thesis. supervised by B. Trousse (AxIS) and K. Avratchenkov (Maestro) and in collaboration with Krisztian Balog (ILPS group, University of Amsterdam).

The goal of expert search is to return a ranked list of people who are knowledgeable on a given topic. Several models have been proposed for expert finding, but so far, these have been focusing solely on returning the most knowledgeable persons as experts on a given topic [70] [101] . In this work we argue that in a real-world organizational setting, the notion of the best expert also depends on the actual user and his needs. We propose a user-oriented approach that balances between two factors that influence the user's choice: time to contact an expert and the knowledge value gained after. We use the distance between the user and an expert in a social network as an indication of contact time, and consider various social graphs, based on organizational hierarchy, geographical location, and collaboration, as well as the combination of these. We performed evaluation against a state-of-the-art baseline on the UvT Expert Collection using graded relevance judgements collected from real users, and demonstrated that the user-oriented approach significantly and substantially outperforms the baseline for all retrieval measures, namely, Mean Average Precision, Mean Reciprocal Rank, and precision at rank 5.

This work has been accepted to ECIR 2011 (only 20% of submissions were accepted)..