|
Title:
|
USING THE TIME COURSE OF SOCIAL BOOKMARKING TO ENHANCE SEARCH |
|
Author(s):
|
Taiki Ueno, Michiaki Yasumura |
|
ISBN:
|
978-989-8533-01-2 |
|
Editors:
|
Bebo White, Pedro Isaías and Flávia Maria Santoro |
|
Year:
|
2011 |
|
Edition:
|
Single |
|
Keywords:
|
Web Search, Information Filtering, Social Bookmarking, Time course of Social Data, Time-based Search |
|
Type:
|
Poster/Demonstration |
|
First Page:
|
609 |
|
Last Page:
|
612 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
The ever increasing size and heterogeneity of the Web is challenging the ability of search engines to find the most relevant material in amongst thousands or millions of hits. With progress on the semantic web stalling, and semantic searching remaining an unsolved problem attempts have been made to use social media and crowdsourcing to improve search results. In this research, we used social bookmarking data to filter data based on how long the information was used for. After analyzing social bookmark data, we found that Web pages that were bookmarked temporarily tend to have less long-term relevance, in contrast to Web pages that were continuously bookmarked for long periods of time tend to have information useful for long periods of time. Based on this observation consulting the utility of temporal social bookmarking data for inferring relevance, we developed a Web service called SelectBookmark that selects content that has been utilized for long periods of time, searching over a huge variety of information. We also report on an initial evaluation of SelectBookmark in terms of whether or not it could find and gather Web pages with higher precision than existing methods of search. |
|
|
|
|
|
|
