Welcome!
To use the personalized features of this site, please log in or register.
If you have forgotten your username or password, we can help.
My Menu
Saved Items

Collaborative web search: a robustness analysis

Michael P. O’MahonyContact Information and Barry SmythContact Information

(1)  School of Computer Science and Informatics, University College Dublin, Dublin, Ireland

Received: 14 January 2008  Accepted: 14 January 2008  Published online: 13 September 2008

Abstract  Collaborative web search utilises past search histories in a community of like-minded users to improve the quality of search results. Search results that have been selected by community members for past queries are promoted in response to similar queries that occur in the future. The I-SPY system is one example of such a collaborative approach to search. As is the case with all open systems, however, it is difficult to establish the integrity of those who access a system and thus the potential for malicious attack exists. In this paper we investigate the robustness of the I-SPY system to attack. In particular, we consider attack scenarios whereby malicious agents seek to promote particular result pages within a community. In addition, we analyse robustness in the context of community homogeneity, and we show that this key characteristic of communities has implications for system robustness.

Keywords  Collaborative web search - Personalisation - Malicious attack - Robustness


Contact Information Michael P. O’Mahony (Corresponding author)
Email: michael.p.omahony@ucd.ie

Contact Information Barry Smyth
Email: barry.smyth@ucd.ie
Fulltext Preview (Small, Large)
Image of the first page of the fulltext

References secured to subscribers.



Export this article
Export this article as RIS | Text
 
Remote Address: 38.107.191.110 • Server: MPWEB26
HTTP User Agent: CCBot/1.0 (+http://www.commoncrawl.org/bot.html)