Abstract
This paper describes the details of our participation in expert search task of the TREC 2007 Enterprise track. The presented study demonstrates the predicting potential of the expertise evidence that can be found outside of
the organization. We discovered that combining the ranking built solely on the Enterprise data with the Global Web
based ranking may produce significant increases in performance. However, our main goal was to explore whether
this result can be further improved by using various quality measures to distinguish among web result items. While,
indeed, it was beneficial to use some of these measures, especially those measuring relevance of URL strings and titles,
it stayed unclear whether they are decisively important.
Original language | Undefined |
---|---|
Title of host publication | Proceedings of the seventeenth Text REtrieval Conference (TREC 2008) |
Place of Publication | Gaithersburg, USA |
Publisher | National Institute of Standards and Technology |
Pages | 52 |
Number of pages | 4 |
ISBN (Print) | not assigned |
Publication status | Published - Feb 2009 |
Event | Seventeenth Text REtrieval Conference, TREC-17 2008 - Gaithersburg, United States Duration: 18 Nov 2008 → 21 Nov 2008 Conference number: 17 |
Publication series
Name | NIST Special Publications |
---|---|
Publisher | USA National Institute of Standards and Technology |
Number | SP 500-277 |
Conference
Conference | Seventeenth Text REtrieval Conference, TREC-17 2008 |
---|---|
Abbreviated title | TREC |
Country/Territory | United States |
City | Gaithersburg |
Period | 18/11/08 → 21/11/08 |
Keywords
- METIS-264417
- EWI-15898
- Expert Search
- IR-67570