Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form
D-Scholarship @ Pitt Banner and Links to Homepage

USER CONTROLLABILITY IN A HYBRID RECOMMENDER SYSTEM

Parra, Denis (2013) USER CONTROLLABILITY IN A HYBRID RECOMMENDER SYSTEM. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

[img]
Preview
 PDF (PhD Dissertation: USER CONTROLLABILITY IN A HYBRID RECOMMENDER SYSTEM)
Primary Text
Download (3MB) | Preview

Abstract

Since the introduction of Tapestry in 1990, research on recommender systems has traditionally focused on the development of algorithms whose goal is to increase the accuracy of predicting users’ taste based on historical data. In the last decade, this research has diversified, with human factors being one area that has received increased attention. Users’ characteristics, such as trusting propensity and interest in a domain, or systems’ characteristics, such as explainability and transparency, have been shown to have an effect on improving the user experience with a recommender. This dissertation investigates on the role of controllability and user characteristics upon the engagement and experience of users of a hybrid recommender system. A hybrid recommender is a system that integrates the results of different algorithms to produce a single set of recommendations. This research examines whether allowing the user to control the process of fusing or integrating different algorithms (i.e., different sources of relevance) results in increased engagement and a better user experience. The essential contribution of this dissertation is an extensive study of controllability in a hybrid fusion scenario. In particular, the introduction of an interactive Venn diagram visualization, combined with sliders explored in a previous work, can provide an efficient visual paradigm for information filtering with a hybrid recommender that fuses different prospects of relevance with overlapping recommended items. This dissertation also provides a three-fold evaluation of the user experience: objective metrics, subjective user perception, and behavioral measures.

Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Parra, Denisdap89@pitt.eduDAP89
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Thesis AdvisorBrusilovsky, Peterpeterb@pitt.eduPETERB
Committee MemberFarzan, Rostarfarzan@pitt.eduRFARZAN
Committee MemberHirtle, Stephen C.hirtle@pitt.eduHIRTLE
Committee MemberSpring, Michael B
Committee MemberResnick, Paulpresnick@umich.edu
Date: 28 August 2013
Date Type: Publication
Defense Date: 22 July 2013
Approval Date: 28 August 2013
Submission Date: 28 August 2013
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 194
Institution: University of Pittsburgh
Schools and Programs: School of Information Sciences > Information Science
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Recommender Systems, User Studies, Information Visualization, Hybrid Recommender
Date Deposited: 28 Aug 2013 12:44
Last Modified: 15 Nov 2016 14:15
URI: http://d-scholarship.pitt.edu/id/eprint/19733

Metrics

Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item