Abstract
This Paper presents the work done for the TREC 2010 entity track. We concentrate on constructing enriched anchor text model by exploiting hierarchical information presented in web pages to retrieve promising pages, and heuristic rules to extract potential candidate entities by zooming in the right section.
Original language | English |
---|---|
Journal | NIST Special Publication |
Publication status | Published - 2010 |
Event | 19th Text REtrieval Conference, TREC 2010 - Gaithersburg, MD, United States Duration: 16 Nov 2010 → 19 Nov 2010 |