|Concordia University at the TREC 2007 QA track|
|Author(s)||Razmara M., Fee A., Kosseim L.|
|Published in||NIST Special Publication|
|Keyword(s)||Unknown (Extra: Clustering methods, Concordia University, Document collection, Factoid questions, QA tracks, Question Answering track, Wikipedia, Information retrieval)|
|Article||BASE, CiteSeerX, Google Scholar|
|Web||Ask, Bing, Google (PDF), Yahoo!|
|Download and mirrors|
|Local copy||Not available|
|Remote mirror(s)||Not available|
|Export and share|
|BibTeX, CSV, RDF, JSON|
|Browse properties · List of conference papers|
In this paper, we describe the system we used for the TREC-2007 Question Answering Track. For factoid questions our redundancy-based approach using a modified version of ARANEA was enhanced further. Our list question answerer uses a clustering method to group the candidate answers that co-occur together. It also uses the target and question keywords as spies to pinpoint the right cluster of candidates. To answer other types of questions, our system extracts from Wikipedia articles a list of interest-marking terms and uses them to extract and score sentences from the AQUAINT-2 and BLOG document collections using various interest-marking triggers.
- This section requires expansion. Please, help!
Probably, this publication is cited by others, but there are no articles available for them in WikiPapers.