March 25 2008, 15.00 hrs

Pavel Serdyukov

Abstract:

In my talk I will address the problem of searching for knowledgeable persons within the enterprise, known as the expert finding (or expert search) task. I will present a probabilistic model using the assumption that terms in retrieved documents are produced by people who are mentioned in them. In other words, this model regards documents as mixtures of candidate experts language models. I will present two ways of personal language models extraction, as well as the way of combining them with other evidences of expertise. I will show that experiments conducted with the TREC Enterprise collection demonstrate the superiority of our approach in comparison with the best one among existing solutions.

	Home > ... > Past colloquia > March 25 2008, 15.00 hrs
<< EWI Home Staff Research Teaching Publications Intranet Colloquia DB Vacancies News Search	March 25 2008, 15.00 hrs Pavel Serdyukov Abstract: In my talk I will address the problem of searching for knowledgeable persons within the enterprise, known as the expert finding (or expert search) task. I will present a probabilistic model using the assumption that terms in retrieved documents are produced by people who are mentioned in them. In other words, this model regards documents as mixtures of candidate experts language models. I will present two ways of personal language models extraction, as well as the way of combining them with other evidences of expertise. I will show that experiments conducted with the TREC Enterprise collection demonstrate the superiority of our approach in comparison with the best one among existing solutions.