AWARD NUMBER: 0546443

INSTITUTION: University of Minnesota, Twin Cities

NSF PROGRAM: Information and Knowledge Management

PRINCIPAL INVESTIGATOR: Gediminas Adomavicius

PROJECT TITLE: CAREER: Next Generation Personalization Technologies

PROJECT SUMMARY: Due to the rapid growth of the Internet and the continuously increasing accessibility to communication technologies and to a vast amount of information on the Web, the problem of information overload has become increasingly more visible in recent years. Various personalization technologies have been proposed to address this problem, including recommender systems that represent the most researched and developed personalization techniques applicable to various types of personalized offerings. Most current-generation recommender systems focus on recommending items to users and represent user preferences for an item with a single rating. Judging by the amount of attention this research area received in the last decade, this approach has worked well in certain applications. However, the inherent limitations of the recommendation framework are preventing recommender systems from tackling more complex personalization applications. In such applications, representing potentially complex user preferences with a single rating and ignoring the contextual information of personalization process is likely to negatively affect the predictive performance of recommender systems. We propose to develop an enhanced recommendation framework by introducing the ideas of context awareness, multi-criteria ratings, rating aggregation, recommendation flexibility, and non-intrusiveness into the recommendation process. Based on these ideas, the newly developed techniques will provide better personalization quality.

PROJECT DATES: June 1, 2006 - May 31, 2012

CURRENT AND FORMER STUDENTS:

• Jesse Bockstedt (graduate of Ph.D. program, Information and Decision Sciences -- currently a faculty member at University of Arizona)
• YoungOk Kwon (graduate of Ph.D. program, Information and Decision Sciences -- currently a faculty member in Sookmyung Women's University, Seoul, Korea)
• Jingjing Zhang (graduate of Ph.D. program, Information and Decision Sciences -- currently a faculty member at Indiana University)
• Rimpi Gupta (graduate of M.S. program, Computer Science)
• Sreeharsha Kamireddy (graduate of M.S. program, Computer Science)
• Praveen Kannan (graduate of M.S. program, Computer Science)
• Vishnu Parimi (graduate of M.S. program, Computer Science)
• Pavithra Ramakrishnan (graduate of M.S. program, Computer Science)
• Sairam Krishnamurthy (graduate of M.S. program, Computer Science)

Project 1: Multi-criteria recommender systems

Publications:

• G. Adomavicius and Y. Kwon. "New Recommendation Techniques for Multi-Criteria Rating Systems." IEEE Intelligent Systems, vol. 22, no. 3. Working paper version can be found here.
• G. Adomavicius, N. Manouselis, and Y. Kwon. "Multi-Criteria Recommender Systems" [book chapter]. In Recommender Systems Handbook: A Complete Guide for Research Scientists and Practitioners, Springer, 2010. Preliminary version can be found here.
• G. Adomavicius and Y. Kwon. "Toward More Accurate and Diverse Recommendations: Integrating Ranking-Based Approaches with Multi-Criteria Rating Information." Working paper. Paper is available upon request.

Project 2: Recommendation query language

Publications:

• G. Adomavicius, A. Tuzhilin, and R. Zheng. "REQUEST: A Query Language for Customizing Recommendations." Information Systems Research, 22(1):99-117, 2011. Working paper version can be found here.

Project 3: Diversity in recommender systems

Publications:

• G. Adomavicius, S. Kamireddy, and Y. Kwon. "Towards More Confident Recommendations: Improving Recommender Systems Using Filtering Approach Based on Rating Variance." Proceedings of the 17th Workshop on Information Technology and Systems (WITS’07), Montreal, Canada, December 2007. Working paper version can be found here.
• G. Adomavicius and Y. Kwon. "Exploring the Effects of Rating Variance in Recommender Systems" [abstract only]. INFORMS Annual Meeting , Washington, D.C., October 2008.
• Y. Kwon. "Improving Top-N Recommendation Techniques Using Rating Variance" [doctoral consortium]. Proceedings of the 2008 ACM Conference on Recommender Systems (RecSys'08), Lausanne, Switzerland, October 2008. Paper can be found here.
• G. Adomavicius and Y. Kwon. "Overcoming Accuracy-Diversity Tradeoff in Recommender Systems: A Variance-Based Approach." Proceedings of the 18th Workshop on Information Technology and Systems (WITS’08), Paris, France, December 2008. Nominated for the Best Paper Award. Working paper version can be found here.
• G. Adomavicius and Y. Kwon. "Toward More Diverse Recommendations: Item Re-Ranking Methods for Recommender Systems." Proceedings of the 19th Workshop on Information Technology and Systems (WITS’09), Phoenix, Arizona, December 2009. Nominated for the Best Paper Award. Won the Best Student Paper Award. Working paper version can be found here. Also presented at:
• G. Adomavicius and Y. Kwon. "Improving Recommendation Diversity: A Ranking-Based Approach." 'Marketing Meets Data Mining' Conference , Austin, Texas, August 2009 [abstract only].
• G. Adomavicius and Y. Kwon. "Techniques for More Accurate and Diverse Recommendations." INFORMS Annual Meeting, San Diego, California, October 2009 [abstract only].
• G. Adomavicius and Y. Kwon. "Improving Aggregate Recommendation Diversity Using Ranking-Based Techniques." IEEE Transactions on Knowledge and Data Engineering, 24(5):896-911, 2012. Official version can be found here.
• G. Adomavicius and Y. Kwon. "Maximizing Aggregate Recommendation Diversity: A Graph-Theoretic Approach." ACM RecSys 2011 International Workshop on Novelty and Diversity in Recommender Systems (DiveRS 2011), Chicago, Illinois, October 2011. Paper can be found here.
• G. Adomavicius and Y. Kwon. "Optimization-Based Approaches for Maximizing Aggregate Recommendation Diversity." INFORMS Journal on Computing. Forthcoming. Paper is available upon request.

Project 4: Clustering temporal data

Publications:

• G. Adomavicius and J. Bockstedt. "C-TREND: Temporal Cluster Graphs for Identifying and Visualizing Trends in Multi-Attribute Transactional Data." IEEE Transactions on Knowledge and Data Engineering, vol. 20, no. 6. Working paper version can be found here.
• J. Bockstedt and G. Adomavicius. "A Visual Mapping Approach for Trend Identification in Multi-Attribute Data." Proceedings of the 17th Workshop on Information Technology and Systems (WITS’07), Montreal, Canada, December 2007. Working paper version can be found here.
• G. Adomavicius, J. Bockstedt, and V. Parimi. "Clustering Large Datasets: New Techniques and Insights" [abstract only]. INFORMS Annual Meeting , Washington, D.C., October 2008.
• G. Adomavicius, J. Bockstedt, and V. Parimi. "Scalable Temporal Clustering for Massive Multidimensional Data Streams." Proceedings of the 18th Workshop on Information Technology and Systems (WITS’08), Paris, France, December 2008. Working paper version can be found here.
• G. Adomavicius, J. Bockstedt, and V. Parimi. "Scalable Clustering of Massive Datasets: A Merge-and-Split Approach for Intelligent Cluster Reconstruction." 2011 Winter Conference on Business Intelligence, Salt Lake City, Utah, March 2011. Paper is available upon request.

Project 5: Context-aware recommender systems

Publications:

• G. Adomavicius and A. Tuzhilin. "Context-Aware Recommender Systems" [tutorial]. Proceedings of the 2008 ACM Conference on Recommender Systems (RecSys'08), Lausanne, Switzerland, October 2008. Tutorial slides can be found here.
• G. Adomavicius and A. Tuzhilin. "Context-Aware Recommender Systems" [book chapter]. Recommender Systems Handbook: A Complete Guide for Research Scientists and Practitioners, Springer, 2010. Forthcoming. Preliminary version can be found here.
• G. Adomavicius, B. Mobasher, F. Ricci, and A. Tuzhilin. "Context-Aware Recommender Systems." AI Magazine, 32(3):67-80, 2011. Paper is available here.
• G. Adomavicius and D. Jannach. "Preface to the Special Issue on Context-Aware Recommender Systems." User Modeling and User-Adapted Interaction. Forthcoming. Paper is available here.

Other related activities:

• Co-Chair, Workshop on Context-Aware Recommender Systems (CARS) : CARS-2009, CARS-2010, CARS-2011, CARS-2012.
• Co-Editor, User Modeling and User-Adapted Interaction special issue on "Context-Aware Recommender Systems"

Project 6: Impact of recommender systems on user preferences

Publications:

• G. Adomavicius, J. Bockstedt, S. Curley, and J. Zhang. "Recommender Systems, Consumer Preferences, and Anchoring Effects." ACM RecSys 2011 Workshop on Human Decision Making in Recommender Systems (Decisions@RecSys 201), Chicago, Illinois, October 2011. Paper is available here. Also presented at:
• 2010 Winter Conference on Business Intelligence, Salt Lake City, Utah, March 2010.
• G. Adomavicius, J. Bockstedt, S. Curley, and J. Zhang. "Do Recommender Systems Manipulate Consumer Preferences? A Study of Anchoring Effects" (extended version). Information Systems Research (conditionally accepted). Preliminary version is available here.
• G. Adomavicius, J. Bockstedt, S. Curley, and J. Zhang. "Effects of Online Recommendations on Consumers’ Willingness to Pay." ACM RecSys 2012 Workshop on Human Decision Making in Recommender Systems (Decisions@RecSys 2012), Dublin, Ireland, September 2012. Paper is available here. Also presented at:
• INFORMS Conference on Information Systems and Technology (CIST 2012), Phoenix, Arizona, October 2012.
• 2012 North American Economic Science Association (ESA) Conference, Tucson, Arizona, November 2012.
• 2013 Winter Conference on Business Intelligence, Snowbird, Utah, March 2013.

Project 7: Stability of recommendation algorithms

Publications:

• G. Adomavicius and J. Zhang. "On the Stability of Recommendation Algorithms." Proceedings of the 4th ACM Conference on Recommender Systems (RecSys 2010), Barcelona, Spain, September 2010. Paper is available here. Also presented at:
• INFORMS Annual Meeting, Austin, Texas, November 2010 [abstract only].
• G. Adomavicius and J. Zhang. "Stability of Recommendation Algorithms." ACM Transactions on Information Systems, 30(4), Article 23 (31 pages), November 2012. Paper is available here.
• G. Adomavicius and J. Zhang. "Maximizing Stability of Recommendation Algorithms: A Collective Inference Approach." 21st Workshop on Information Technology and Systems (WITS 2011), Shanghai, China, December 2011. Winner of the Best Paper Award. Paper is available here. Also presented at:
• INFORMS Annual Meeting, Charlotte, North Carolina, November 2011 [abstract only].
• 2012 Winter Conference on Business Intelligence, Snowbird, Utah, March 2012.
• "Iterative Smoothing Technique for Improving Stability of Recommender Systems." ACM RecSys 2012 Workshop on Recommender Utility Evaluation: Beyond RMSE (RUE 2012), Dublin, Ireland, September 2012. The paper can found here.
• G. Adomavicius and J. Zhang. "Classification and Ranking Stability of Recommendation Algorithms." Working Paper. Paper is available upon request.

Project 8: Impact of data characteristics on recommender systems performance

Publications:

• G. Adomavicius, Y. Kwon, and J. Zhang. "Impact of Data Characteristics on Recommender Systems Performance." Proceedings of the 20th Workshop on Information Technology and Systems (WITS’10), St. Louis, Missouri, December 2010. Paper is available here.
• G. Adomavicius and J. Zhang. "Impact of Data Characteristics on Recommender Systems Performance" (extended version). ACM Transactions on Management Information Systems, 3(1), Article 3 (17 pages), April 2012. The paper can found here.

Last modified: May 1, 2013