Long query as an effective method for improving the quality of information retrieval on the Web.
Item
-
Title
-
Long query as an effective method for improving the quality of information retrieval on the Web.
-
Identifier
-
AAI3037447
-
identifier
-
3037447
-
Creator
-
Taksa, Isak.
-
Contributor
-
Adviser: Jacob Shapiro
-
Date
-
2002
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Computer Science
-
Abstract
-
The quality of information retrieval on the Web is a subject of great interest to the Information Retrieval community as well as to a very large and growing Community of Web users. This thesis addresses how long queries can improve the quality of information retrieval on the Web. Long queries allow the user to write the query in natural language, making it easier for the user to completely describe the information need. Using long queries in the Web environment has not been addressed in its entirety in the scientific literature nor has it been adequately incorporated into any commercial search engine.;We propose and analyze several novel algorithms dealing with all aspects of long query information retrieval on the Web such as selecting of search terms and phrases, constructing multiple query formulations, merging and ranking search results. We also present a user feedback option to direct the query formulation process and a new user interface especially designed to handle long queries.;We developed a meta-search engine, which incorporated the proposed algorithms, the feedback option, and the user interface. We performed a series of experiments to evaluate the new meta-search engine and compare its search results with the results of popular search engines on the Web. These experiments clearly demonstrated that using long queries in the Web environment is practical and can substantially improve the quality of information retrieval.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
