Haidong Wang

Graduation Semester and Year

2008

Language

English

Document Type

Thesis

Degree Name

Master of Science in Computer Science

Department

Computer Science and Engineering

First Advisor

Gautam Das

Abstract

In this thesis, we propose minimum-effort driven navigational techniques for enterprise database systems based on the faceted search paradigm. Our proposed techniques dynamically suggest facets for drilling down into the database such that the cost of navigation is minimized. At every step, the system asks the user a question or a set of questions on different facets and depending on the user response, dynamically fetches the next most promising set of facets, and the process repeats. Facets are selected based on their ability to rapidly drill down to the most promising tuples, as well as on the ability of the user to provide desired values for them. Our facet selection algorithms also work in conjunction with any ranked retrieval model where a ranking function imposes a bias over the user preferences for the selected tuples. Our methods are principled as well as efficient, and our experimental study validates their effectiveness on several application scenarios.

Disciplines

Computer Sciences | Physical Sciences and Mathematics

Comments

Degree granted by The University of Texas at Arlington

Download

Share

COinS