|
Title:
|
SEARCHING BY COMMON TERMS - A MODEL AND IDEAL RANKING METHOD FOR NOVICE SEARCHERS |
|
Author(s):
|
Izumi Suzuki , Yoshiki Mikami , Ario Ohsato |
|
ISBN:
|
978-972-8924-69-0 |
|
Editors:
|
Kinshuk, Demetrios G Sampson, J. Michael Spector, Pedro Isaías and Dirk Ifenthaler |
|
Year:
|
2008 |
|
Edition:
|
Single |
|
Keywords:
|
Searching behavior, user model, query term, relevance ranking, search engines, single-meaning terms |
|
Type:
|
Reflection Paper |
|
First Page:
|
405 |
|
Last Page:
|
407 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
Searching by common terms, which has been considered a difficult process, is modeled, and a relevance ranking method
that uses shorten queries is introduced in this paper. The search model assumes that a query contains many (more than 3)
terms. First, 10 or 20 queries are generated by removing unimportant terms from the original query. Term importance is
assumed to be given by the user. Next, searches are conducted with these queries, and the search results are analyzed to
re-rank the pages. In an experiment with 16 examples, this ranking method attained better performance than that of the
user model. |
|
|
|
|
|
|
