Topic Detection and Tracking in Personal Search History

Gupta, Kamal Kant

Permalink

https://hdl.handle.net/2142/11470 Copy

Description

Title

Topic Detection and Tracking in Personal Search History

Author(s)

Gupta, Kamal Kant

Issue Date

2008-07

Keyword(s)

computer science

Abstract

This thesis describes a system for tracking and detecting topics in personal search history. In particular, we developed a time tracking tool that helps users in analyzing their time and discovering their activity patterns. The system allows a user to specify interesting topics to monitor with a keyword description. The system would then keep track of the log and the time spent on each document and produce a time graph to show how much time has been spent on each topic to be monitored. The system can also detect new topics and potentially recommend relevant information about them to the user. This work has been integrated with the UCAIR Toolbar, a client side agent. Considering limited resources on the client side, we designed an e±cient incremental algorithm for topic tracking and detection. Various unsupervised learning approaches have been considered to improve the accuracy in categorizing the user log into appropriate categories. Experiments show that our tool is effective in categorizing the documents into existing categories and detecting the new useful catgeories. Moreover, the quality of categorization improves over time as more and more log is available.

Type of Resource

text

Permalink

http://hdl.handle.net/2142/11470

Copyright and License Information

You are granted permission for the non-commercial reproduction, distribution, display, and performance of this technical report in any format, BUT this permission is only for a period of 45 (forty-five) days from the most recent time that you verified that this technical report is still available from the University of Illinois at Urbana-Champaign Computer Science Department under terms that include this permission. All other rights are reserved by the author(s).

Owning Collections