Scott Buffett, PhD (UNB)
Associate Research Officer
Learning and Collaborative Technologies Group
Information and Communications Technology Portfolio
National Research Council Canada

Adjunct Professor
Faculty of Computer Science
University of New Brunswick
scott.buffett@nrc.gc.ca
Research

My primary area of research lies in artificial intelligence, focussing on both advancing the state of the art in AI, as well as seeking opportunities for applying AI to other problems. Generally my primary interests lie in complex decision making, where often the theory of utility is employed to assist an actor to make optimal choices in the face of uncertainty or excessive information/data. Some areas where this is relevant:

Multi-agent systems: Many problem areas involving multi-agent systems exist, from automated negotiation to collaborative problem solving to resource allocation. My work in these areas looks specifically at how agents can interact in self-interested ways that will result in solutions for the common good. One example of this is in the development of mechanisms for automated negotiation of transactions between buyers and sellers that are influenced by supply and demand. Agents act rationally in terms of maximizing their own utility, but at the same time social welfare of the entire society is increased. I have developed the OmniBid system (see youtube video here, beginning around 4 minute mark) as a working embodiment of this concept.

Preference Elicitation: In order to utilize multi-agent systems in a practical setting, where perhaps agents are used to make decisions on behalf of a human user, techniques are needed to extract the necessary information on the users in order for the agents to act optimally. While these techniques must not be overly cumbersome or intrusive, they generally must be capable of acquiring a large amount of information in order to be effective. My work in this area focuses on preference and utility prediction based on limit user preference data.

Workflow/Process mining: Process mining is an area of data analytics that attempts to identify patterns in event data generated by machines, computers, etc., and to automatically build process models that depict or represent the typical workflows of activity that take place on those machines. Thus process mining can help a company to extract a holistic view of their operational workflows from the reams of data generated from various processes, allowing them to easily identify inefficiencies, bottlenecks, best practices, etc. My work in this area involves the creation of a new theory of process mining that shifts the focus from the more passive act of mining data for existing processes, to the more active approach of dynamically identifying workflows for processes to be executed. This takes the application of process mining from mere business analytics to a world where, say, equipment operators in an oil mining or manufacturing operation can gain real-time guidance on how to proceed in such a way as to ensure best practices are met or safety regulations are followed, to name a few. The attractiveness of this approach is that the dynamic composition of these procedures is entirely data driven, as opposed to the more traditional method of creating and locating such procedures in an outdated and unused binder on the shelf. This new direction, which is made possible with the use of techniques from artificial intelligence such as logical reasoning and planning, is particularly challenging due to the fact that there may not exist historical data for exactly accomplishing the task at hand in the exact same situation. Thus a host of new research opportunities exist in this area.

Electronic/Social Commerce: In the past, a major focus of mine was in the area of electronic commerce as an application of my work in AI. In recent years, this focus has shifted to the area of "social commerce", which focuses on how people do commerce in the context of the social connectivity that exists today. Techniques from social network analysis have thus become critical to understanding and exploiting social commerce as a research field and advancing the state of the art. My contributions to the field have centered on analysis and simulation of social networks and how they affect commerce.

I'm also involved in work in privacy, data analytics for energy management, data mining, machine learning and Petri nets. Before diving into the world of agents and e-commerce for my PhD work, I did my Masters in automated theorem proving.


Grad Students


Teaching

Current Courses:

Past Courses:

Publications

See my NRC publications page to download papers

  • Lee, K. H., Buffett, S., Fleming, M. W. Maintaining Preference Networks That Adapt to Changing Preferences. In Proceedings of the Canadian Conference on Artificial Intelligence (CAI 2013), Regina, Canada, May 2013, pp. 89-99.
  • Spencer, B., Buffett, S. Simulating social commerce applied to buyer group pricing, recommendation incentives, and bundling. In Proceedings of the 14th Annual International Conference on Electronic Commerce (ICEC 2012), Singapore, Aug 2012, pp. 95-98.
  • Bediako-Asare, H., Buffett, S., and Fleming, M. W. Utility Estimation in Large Preference Graphs Using A* Search. In Proc. of the Canadian Conference on Artificial Intelligence (CAI 2011), Lecture Notes in Computer Science, 2011, Volume 6657/2011, 50-55.
  • Buffett, S., and Geng, L. Using Classification Methods to Label Tasks in Process Mining. Journal of Software Maintenance and Evolution: Research and Practice, 22:6-7, Summer 2010, Pages 497-517.
  • Buffett, S. A Revelation Mechanism for Shared Conditional Preferences in Multi-Attribute Negotiation. The Third International Workshop on Agent-based Complex Automated Negotiations, held in conjunction with the Eighth International Joint Conference on Autonomous Agents and Multi-Agent Systems (AAMAS 2010), Toronto, Canada, 10-14 May 2010, pages 9-15.
  • Geng, L., Buffett, S., Hamilton, B., Wang, X., Korba, L., Liu, H., Wang, Y. Discovering Structured Event Logs from Unstructured Audit Trails for Workflow Mining.18th International Symposium on Methodologies for Intelligent Systems, Prague, Czech Republic, Sept 2009.
  • Buffett, S. Abductive Workflow Mining using Binary Resolution on Task Successor Rules. The International RuleML Symposium on Rule Interchange and Applications (RuleML 2008). October 30, 2008. NRC 50392.
  • Buffett, S., Hamilton, B. Abductive Workflow Mining. 4th Workshop on Business Process Intelligence (BPI 08) to be held in conjunction with Business Process Management (BPM 2008). September 1, 2008. NRC 50393.
  • Buffett, S., and Geng, L. Bayesian Classification of Events for Task Labeling Using Workflow Models. 4th Workshop on Business Process Intelligence (BPI 08) to be held in conjunction with Business Process Management (BPM 2008). September 1, 2008. NRC 50389
  • Korba, L., Wang, Y., Geng, L., Song, R., Yee, G., Patrick, A.S., Buffett, S., Liu, H., and You, Y. (2008). Private data discovery for privacy compliance in collaborative environments. Proceedings of the Fifth International Conference on Cooperative Design, Visualization and Engineering (CDVE 2008). Palma de Mallorca, Mallorca. September 21-25, 2008. NRC 50386
  • Qin, M., Buffett, S., and Fleming, M.W. Predicting User Preferences via Similarity-Based Clustering. Canadian Artificial Intelligence Conference (AI 2008), Windsor, Ontario. May 27 - 30, 2008. NRC 50333.
  • Buffett, S., and Spencer, B. A Bayesian Classifier for Learning Opponents' Preferences in Multi-Object Automated Negotiation. Electronic Commerce Research and Applications Journal. 2006. NRC 48500. Volume 6, Issue 3, Autumn 2007, Pages 274-284
  • Chen, S., Buffett, S., and Fleming, M. "Reasoning with Conditional Preferences across Attributes," The 20th Canadian Conference on Artificial Intelligence May 28, 2007 NRC 49292.
  • Buffett, S., and Fleming, M.W. Applying a Preference Modeling Structure to User Privacy. Workshop on Sustaining Privacy in Autonomous Collaborative Environments (SPACE 2007), held in conjunction with the IFIP Conference on Trust Management (IFIPTM07) Moncton, NB July 30, 2007. NRC 49372.
  • L. Korba, R. Song, G. Yee, A. Patrick, S. Buffett, Y. Wang, L. Geng. Private Data Management in Collaborative Environments. The Fourth International Conference on Cooperative Design, Visualization and Engineering (CDVE2007), Shanghai, PR China , Sept. 16-20, 2007.
  • Buffett, S., and Fleming, M.W. Persistently Effective Query Selection in Preference Elicitation. The 2007 IEEE/WIC/ACM International Conference on Intelligent Agent Technology (IAT'07). Fremont, CA. November 2-5, 2007. NRC 49846.
  • Buffett, S., Kosa, T.A. "Towards a Model for Risk and Consent Management of Private Health Information," The Conference on Privacy, Security and Trust (PST2006). Toronto, Ontario, Canada. October 31, 2006. NRC 48746.
  • Buffett, S., Spencer, B., Comeau, L., Fleming, M.W. "Detecting Opponent Concessions in Multi-Issue Automated Negotiation," The 8th International Conference on Electronic Commerce (ICEC'06). Fredericton, New Brunswick, Canada. August 14, 2006. NRC 48552.
  • Buffett, S., Fasli, M. "TAC-REM - The Real Estate Market Game: A Proposal for the Trading Agent Competition," Trading Agent Design and Analysis and Agent Mediated Electronic Commerce VIII. A workshop held in conjunction with the Fifth International Joint Conference on Autonomous Agents and Multi-Agent System (AAMAS-06). Hakodate, Japan. May 8-12 2006. NRC 48498.
  • Flinn, S., Buffett, S. "Exercising the Right of Privacy," Privacy Protection for E-Services. Idea Group Inc. George Yee (Editor). 2005. NRC 48261.
  • Buffett, S., Spencer, B. "Learning Opponents' Preferences in Multi-Object Automated Negotiation," Seventh International Conference on Electronic Commerce (ICEC?05). Xi'an, China. August 15-17, 2005. NRC 48260.
  • Buffett, S., Comeau, L., Fleming, M.W., Spencer, B. "MONOLOGUE: A Tool for Negotiating Exchanges of Private Information in E-Commerce," Third Annual Conference on Privacy, Security and Trust (PST'05). St. Andrews, New Brunswick, Canada. October 12-14, 2005. NRC 48259.
  • Buffett, S., A Markov Model for Inventory Level Optimization in Supply-Chain Management, The 18th Conference of the Canadian Society for Computational Studies of Intelligence (AI 2005). Victoria, British Columbia, Canada. May 9-11, 2005. To appear.
  • Buffett, S., Keping, J., Liu, S., Spencer, B., Wang, F. Negotiating Exchanges of P3P-Labeled Information for Compensation, Computational Intelligence Journal, 20(4): 663-677. November 2004. NRC 47409.
  • Buffett, S., Fleming, M.W., Richter, M.M., Scott, N., Spencer, B. Determining Internet Users? Values for Private Information, Second Annual Conference on Privacy, Security and Trust (PST?04). Fredericton, New Brunswick, Canada. October 14-15, 2004. pp. 79-88. NRC 47408.
  • Buffett, S. Considering Expected Utility of Future Bidding Options in Bundle Purchasing with Multiple Auction, The Sixth International Conference on Electronic Commerce (ICEC?04). Delft, The Netherlands. October 25-27, 2004. pp. 69-76. NRC 47407.
  • Buffett, S. and Scott, N. An Algorithm for Procurement in Supply Chain Management>. AAMAS-04 Workshop on Trading Agent Design and Analysis, New York, 2004. pp 9-14
  • Buffett, S., Grant, A. A Decision-Theoretic Algorithm for Bundle Purchasing in Multiple Open Ascending-Price Auction, The 17th Conference of the Canadian Society for Computational Studies of Intelligence (AI 2004). London, Ontario, Canada. May 17-19, 2004. pp. 429-433. NRC 47144.
  • Buffett, S., Spencer, B. A Decision Procedure for Bundle Purchasing with Incomplete Information on Future Prices, International Journal of Electronic Commerce, 8(4): 131-144. 2004. NRC 47143.
  • Buffett, S. Monte Carlo Algorithms for Expected Utility Estimation in Dynamic Purchasing, PhD Thesis. Faculty of Computer Science, University of New Brunswick. Fredericton, New Brunswick, Canada. March 2004. 191 pages. NRC 46557.
  • Buffett, S., Keping, J., Liu, S., Spencer, B., Wang, F. Negotiating Exchanges of P3P-Labeled Information for Measurable Benefits, Business Agents and the Semantic Web 2004 (In conjunction with the 2003 Canadian AI Conference). Halifax, Nova Scotia, Canada. June 14, 2003. pp. 25-34. NRC 47092.
  • Buffett, S., Spencer, B. Efficient Monte Carlo Decision Tree Solution in Dynamic Purchasing Environments, The International Conference on Electronic Commerce (ICEC'03). Pittsburgh, Pennsylvania, USA. October 1, 2003. NRC 46489.
  • Buffett, S. Planning and Procurement in Multi-Agent Systems. Novel E-Commerce Applications of Agents workshop at the Canadian AI-2001 conference, June 8, 2001.
  • S. Buffett and B. Spencer. Reducing the Search Space Required in Implicit AND/OR Tree Solution Search. In Proceedings of the APICS Mathematics, Statistics and Computer Science Conference, St. Mary's University, Halifax, NS, Canada, Oct 1998.
  • S. Buffett. Investigating Iterative Deepening in Top Down Automated Reasoning. Master's Thesis, University of New Brunswick, 1998.



    Last revised January 7, 2014