Informs Annual Meeting 2017

SD26

INFORMS Houston – 2017

SD26

3 - Role Identification and Modeling of Online Health Communities” Jiaqi Zhou, C ity University of Hong Kong, 12-200 Lau Ming Wai, Hong Kong, jiaqi.zhou@my.cityu.edu.hk, Qingpeng Zhang, Mingyang Li Online health communities present a platform for users to exchange information and get social support from other users, particularly those suffering from similar diseases. In this study, we focus on the impact of OHCs on people with depression. We combined text mining, social network analysis, and econometric analysis and characterized their roles with topological measures. Finally, we discovered the changing patterns of users’ roles dynamically and found that patients’ emotion could be significantly enhanced by responses of other OHC users.” 4 - Review of Business Strategy Problems in Social Media Analytics Michel Ballings, University of Tennessee, University of Tennessee, Knoxville, TN, 37996, United States, michel.ballings@utk.edu A review of business strategy research is given with an emphasis on building networks and their roles in marketing and sales. Invited: Auctions Invited Session Chair: Benjamin Lubin, Boston University, Boston, MA, 02215, United States, blubin@bu.edu 1 - Probably Approximately Efficient Combinatorial Auctions via Machine Learning A well-known problem in combinatorial auctions is the exponential growth of the value space in goods, burdening both bidders and the auctioneer. We introduce a new design paradigm for CAs based on machine learning. Bidders report bids to a proxy agent via a small number of value queries. The proxy then uses an ML algorithm to generalize from those bids to the whole value space, and the efficient allocation is computed over these generalizations. We introduce the concept of probably approximate efficiency (PAE) to measure the efficiency of the new ML- based auctions, and formally show how the generelizability of an ML algorithm relates to the efficiency loss incurred by the corresponding ML-based auction. 2 - Efficient Dynamic Auctions for Private Valuations Oleg V. Baranov, University of Colorado-Boulder, 256 UCB, Boulder, CO, 80309, United States, oleg.baranov@colorado.edu The literature on dynamic implementation of the VCG mechanism routinely uses price-demand queries for preference elicitation. As a result, all known dynamic auction procedures that implement the Vickrey outcome for the general private valuation domain are impractical. The paper makes two contributions. First, we show that the general problem of designing a dynamic VCG mechanism is reduced to a problem of specifying an appropriate elicitation process that satisfies certain properties. Second, we characterize an ascending auction that implements the Vickrey outcome for general private valuation domain while using a single linear and anonymous price trajectory for preference elicitation. 3 - Competitive Equilibrium Pricing in Auctions and Exchanges Robert Day, University of Connecticut, 2100 Hillside Road U-1041, Storrs, CT, 06269-1041, United States, Bob.Day@business.uconn.edu Cutting plane methods allow us to generate competitive equilibrium prices for an arbitrary combinatorial auction. We explore the computational aspects of these methods and their extension to the arbitrary case of combinatorial exchanges. Benjamin Lubin, Boston University, School of Management, 595 Commonwealth Avenue, Boston, MA, 02215, United States, blubin@bu.edu, Gianluca Brero, Sven Seuken SD28 350D Combinatorial Mechanisms

350B DEA Applications Invited: Data Envelopment Analysis Invited Session Chair: Kankana Mukherjee, Babson College, Wellesley, MA, 02481, United States, kmukherjee@babson.edu 1 - Efficiency and Productivity Dynamics of Indian Bank Branches Kankana Mukherjee, Babson College, Babson Park, MA, United States, kmukherjee@babson.edu, Subhash C.Ray, Abhiman Das This paper examines the trajectory of efficiency and productivity dynamics of a network of branches of a large Indian bank during the post Great Recession period. We use Data Envelopment Analysis to study branch level data for the years 2009-10 through 2016-17. By isolating the contributions of technical change, efficiency change, and scale change to overall productivity change of each branch, we are able to identify the sources of superior versus inferior branch level performance, which helps to provide guidelines for improving the performance of the bank at the institutional level. 2 - Environmental Productivity Growth in Consumer Durables Xun Zhou, Doctoral Candidate, Aalto University School of Business, Runeberginkatu 22-24, Helsinki, 00100, Finland, xun.zhou@aalto.fi Only one previous study considers environmental performance of consumer durables. This is somewhat surprising because during their use phase consumer durables are actually production units that consume energy to provide services for consumers and thus contributors to pollutions. This study aims to develop an index of environmental productivity change specific to consumer durables. We first specify how to model the production activity during the use of consumer durables. Then we present an overview of conventional measurements of environmental productivity and describe how they can be applied in this study. Finally we use a Finnish dataset to illustrate interpretations of the proposed index. 3 - Building Models for Efficiency and Effectiveness in the Financial Industry Abbas Attarwala, University of Waterloo, 200 University Avenue West, Waterloo, ON, N2L.3G1, Canada, abbas.attarwala@gmail.com, Stanko Dimitrov We review recent work on evaluating models of efficiency and effectiveness for the financial industry. Using the reviewed evaluation procedure and publicly available data we computationally efficiently and systematically search the space of efficiency and effectiveness data envelopment analysis models to present models of efficiency and effectiveness valid for multiple datasets. We conclude with some key managerial takeaways from our findings. 350C Ideas And Discussing Relating to the Future of Social Media Analytics Research Invited: Social Media Analytics Invited Session Chair: Christopher M Smith, Air Force Institute of Technology, Air Force Institute of Technology, Xenia, OH, 45385, United States, cms3am@virginia.edu 1 - Ideas and Discussion Relating to the Future of Social Media Analytics Research Chris Smith, Air Force Institute of Technology, 2950 Hobson Way, 2950 Hobson Way, Wright-Patterson AFB, OH, 45433, United States, cms3am@virginia.edu A brief general overview of our cluster is provided. Trends in the literature are reviewed as well as in topics within past INFORMS clusters. A set of discussion topics are proposed for all the participants and, ideally, for the audience. 2 - Review of Decision Problems in Social Media Analytics Theodore T. Allen, Ohio State University, Columbus, OH, 43210- 1271, United States, allen.515@osu.edu This presentation focuses on reviewing recent decision problems explored in social media analytics. An example is so-called “Timely Decision Analysis” enabled by social media analytics. Yet, many other examples ranging from financial investments to optimal target identification SD27

110