Professor
Joined School 1984
BA (McGill),
MSc (Toronto),
PhD (Toronto)
| rcohen@uwaterloo.ca | |
| Web | http://ai.uwaterloo.ca/~rcohen |
| Voice | 519-888-4567 x34457 |
| Fax | 519-885-1208 |
|
| Robin Cohen
Professor Joined School 1984 BA (McGill),
|
Robin Cohen conducts research in the artificial intelligence subfields of multiagent systems, user modeling and intelligent interaction. Most recently the focus of her research has been on modeling trust in multiagent systems, with connections to social networking and to applications of electronic commerce. She has also been exploring trust modeling for the application of transportation (in mobile vehicular ad-hoc networks) and in the context of peer-based intelligent tutoring systems.
With Jie Zhang, she explored scenarios where buying agents make use of advice provided by other buying agents in order to select the most appropriate selling agents. Included in this research is an approach that allows the combination of private and public modeling of the advisor's trustworthiness. The approach introduces as well incentives for honest reporting from advisors, through rewards provided by sellers to those advisors accepted into a large number of social networks of other buying agents.
With Georgia Kastidou, Professor Cohen has examined the design of an effective community structure within multiagent systems, so that agents can participate in multiple communities and communities can be selective about which agents to embrace. Trust and reputation modeling are central to this framework, which studies methods for communities to reason about agents based in part on their reputation in the communities to which they have previously belonged. Incentive mechanisms to encourage agents to be trustworthy within their communities and communities to be truthful in their reporting about agents are also an important element of the design.
With Reid Kerr, Professor Cohen has investigated the promotion of secure electronic marketplaces, in a framework whereby selling agents are required to possess units of trust in order to engage in transactions and risk losing these trust units, whenever they are untrustworthy. This research also explored some of the common vulnerabilities in existing trust and reputation models, leading to a demonstration of how smart cheating agents can prosper even when their trustworthiness is being modeled. This in turn resulted in the design of a valuable testbed for measuring the performance of any trust and reputation modeling system. Most recently, work with Kerr has focused on the critical challenge of collusion in multiagent systems. At present we are developing techniques aimed at recognizing clusters of agents that avoid harm and instead produce benefit to each other.
In dynamic environments with sparse connections, it is beneficial to employ a multi-dimensional trust model, which takes into consideration the role of each agent that provides information, as well as its location, the time of the communication, its known trustworthiness from direct experience and its trustworthiness reflected from majority opinion in the multiagent system. This approach is currently being explored for the application of mobile vehicular ad-hoc networks.
In environments where student learning can be achieved on the basis of the previous experience of peers, the challenge is to determine the appropriate social network to employ -- which peers are most reputable and which learning experiences have been most beneficial, for students bearing some similarity to the current student. This is the topic of current research with John Champaign.
Finally work with Masters student Joshua Gorner is examining how to determine the ideal size of social network, when other agents are brought in to inform an agent making a decision. There is a fundamental tradeoff between having enough agents to reflect the required experience and not having so many agents that less accurate advice is offered. This research is currently being applied to multiagent trust and reputation modeling systems but is of general value to other social networking applications as well.
David R. Cheriton Faculty Fellow, University of Waterloo (2010-2013); Best Paper Award, International Conference on Information Technology and Convergence (ITCS) with C.Cheng, J. Zhang and P.Ho (2010); Award for Excellence in Graduate Supervision, University of Waterloo (2009); Faculty of Mathematics Distinguished Teaching Award, University of Waterloo (2008); Best Paper Award Canadian Semantic Web Symposium with J. Zhang (2006); Best Paper Award Conference on Privacy, Security and Trust (PST) with K. Regan and P. Poupart (2005); Faculty of Mathematics Faculty Fellowship, University of Waterloo (2002-2005)
During her most recent sabbatical, Professor Cohen focused on research conducted with two of her previous PhD students, Thomas Tran at the University of Ottawa and Fei Song at the University of Guelph. With Tran, she investigated trust modeling for the application of mobile vehicular ad-hoc networks. With Song, she focused on topics of information extraction and document classification, injecting an element of user modeling into tasks such as classifying and identifying people mentioned in webpages. Joint research was also conducted with Pinar Yolum and Murat Sensoy of Bogazici University in Turkey, on the topic of trust modeling for the application of service-oriented computing and with former PhD student Michael Fleming on the topic of modeling the cost of bothering users in a decision-theoretic framework for reasoning about interaction.
Her most recent connection to industry has arisen as result of an NSERC Strategic Research Network project known as hSITE, aimed at developing technological enhancements to improve the delivery of healthcare in both hospital and homecare settings. She has collaborated primarily with a team of researchers at the University of Toronto (including from industrial engineering and from nursing), who in turn have connections to industrial partners, interested in exploring sensing and networking solutions for healthcare. Professor Cohen's involvement in this project is primarily focused on reasoning about interaction with medical experts for hospital decision making.
J. Champaign and R. Cohen; An Annotations Approach to Peer Tutoring; Proceedings of 2010 Educational Data Mining Conference; 2010.
G. Kastidou, K. Larson and R. Cohen; Exchanging Reputation Information Between Communities: A Payment Function Approach; Proceedings of IJCAI 2009; 2009.
R. Kerr and R. Cohen; Smart Cheaters Do Prosper: Defeating Trust and Reputation Systems; Proceedings of AAMAS 2009; 2009.
J. Zhang and R. Cohen. Design of a Mechanism for Promoting Honesty in E-Marketplaces. Proceedings of AAAI, 2007.
K. Regan, P. Poupart, and R. Cohen. Bayesian Reputation Modeling in E-Marketplaces Sensitive to Subjectivity, Deception and Change. Proceedings of AAAI, 2006.
R. Cohen, M. Cheng, and M. Fleming. Why Bother About Bother: Is it Worth it to Ask the User? AAAI Fall Symposium on Mixed-initiative Problem-solving Assistants, 2005.
David R. Cheriton School of Computer Science
University of Waterloo
Waterloo, Ontario, Canada N2L 3G1
Tel: 519-888-4567 x33293
Fax: 519-885-1208
Contact | Feedback: cs-webmaster@cs.uwaterloo.ca | David R. Cheriton School of Computer Science | Faculty of Mathematics
Last modified: Monday, 20-Dec-2010 14:30:05 EST
