Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Improving Recommendations Using Watching Networks in a Social Tagging System

Lee, Danielle and Brusilovsky, Peter (2011) Improving Recommendations Using Watching Networks in a Social Tagging System. In: Proceedings of iConference 2011, 8-11 February 2011, Seattle, Washington, USA.

Abstract

This paper aims to examine whether users' watching networks can improve collaborative filtering-based recommendations (CF). Watching networks are established by users upon their perceived usefulness or interests about other users' information collections. The networks do not require mutual agreement between a watching party and a watched party. The typical example of this network is 'following' in Twitter, 'watching' on CiteULike, or 'contacts' on Flickr. Once a user declares that 'I want to watch user A', the user A's information collection is displayed to the watching user, continuously. It can beinterpreted to mean that a watching user found some shared interests in user A's collection and want to refer to it in future. The approaches explored in this paper take advantage of this watching network as a part of user's preferences for recommendations. To evaluate the potential of these approaches, we focus on a social tagging system, CiteULike. Our data shows that in this context, a hybrid recommendation approachthat fusesCF and watching network-based recommendations outperforms both CF and network-based recommendations.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: Conference or Workshop Item (Paper)
Status: Published
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Lee, Daniellehyl12@pitt.eduHYL12
Brusilovsky, Peterpeterb@pitt.eduPETERB0000-0002-1902-1464
Date: 2011
Date Type: Publication
Event Title: Proceedings of iConference 2011
Event Dates: 8-11 February 2011
Event Type: Conference
Institution: University of Pittsburgh
Schools and Programs: School of Information Sciences > Information Science
Refereed: Yes
Official URL: http://doi.acm.org/10.1145/1940761.1940766
Related URLs:
Other ID: DOI: 10.1145/1940761.1940766
Additional Information: Access to the full text is subject to the publisher's access restrictions.
Date Deposited: 06 Jul 2011 15:38
Last Modified: 15 Nov 2016 13:35
URI: http://d-scholarship.pitt.edu/id/eprint/5980

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item