Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

MA68

about product popularity and quality. However, implementing such a system might also have unintended consequences, hindering product sales due to the lack of control over WOM volume and content. This study examines how implementing a WOM system (through social media integration) on an e- commerce website affects customer conversion in the two stages of consumer purchase funnel, namely, adding a product to cart and placing an order. 2 - Be Gentle to the Newbies: A Randomized Field Experiment on Negative Feedback in Online Communities Wei Chen, University of Arizona, 1130 East Helen Street, McClelland Hall 430, P.O. Box 210108, Tucson, AZ, United States, Laura Brandimarte, Yinchu Zhu, Dong Jing Negative feedback (e.g., downvotes) in online communities helps to control the quality of contribution but may discourage users from participation, especially for new users. In this paper, we conduct a randomized field experiment to evaluate the causal effect of downvotes on user-generated content. We find that downvotes indeed decrease future user contributions. However, this effect only exists for new users who have less experience. Instead, experienced users increase their contributions after negative feedback. Policy evaluation results using machine learning methods provide design suggestions for downvotes base on the characteristics of contributors. 3 - Impact of the Interplay Between Review Volume and Rating in Digital Platforms on Sales: An Empirical Study Samayita Guha, Temple University, Philadelphia, PA, United States, Naveen Kumar, Subodha Kumar, Joydeep Srivastava Impact of the Interplay between Review Volume and Rating in Digital Platforms on Sales: An Empirical Study. Consumers now increasingly depend on user- generated reviews. Based on the data collected from a digital-platform, we empirically examine the impact of interaction between the number of reviews and the average rating. This study would help us in answering an important and relevant question: What is more important - Volume or Quality? 4 - From Broadcasting to Narrowcasting: Examining User Content Consumption, Generation and Sharing in Social Media Lanfei Shi, University of Maryland College Park, College Park, MD, United States User engagement (UE) through content consumption (CC), content generation (CG) or content sharing (CS) on social media has been widely studied. However, the fundamental question of how the 3 dimensions of UE interact remains largely unanswered. Therefore, we design a large-scale randomized field experiment to investigate the causal effect of CC reduction on CG and subsequent CS on Snapchat. We are among the first to examine the interplay of different dimensions of UE using a randomized field experiment for causal inference. Our findings complement the current literature of user behavior on social sites and provide useful insights into platform design and marketing campaigns on social platforms. n MA68 West Bldg 105C IEEE Transactions on Automation Science and Engineering Invited Session Sponsored: Quality, Statistics and Reliability Sponsored Session Chair: Jingshan Li, The University of WisconsinûMadison Co-Chair: Xiao Liu, University of Arkansas 1 - Prescriptive Modeling and Compensation of In-plane Shape Deformation for 3D Printed Freeform Products Qiang Huang, University of Southern California, Dept of Industrial & Systemes Engineering, 3715 Mcclintock Avenue, Ger 240, Los Angeles, CA, 90089, United States, He Luan The shape accuracy of 3D printed products is a critical issue, especially for freeform products with complex geometries. In addition, shape accuracy modeling has to be transparent to specific designs and 3D printing processes for wider applications. This study fills the gap by establishing a data-driven approach to predict and compensate the in-plane (x-y plane) shape deviations of 3D printed freeform products based on a limited number of simple trial shapes. Experimental investigation using stereolithography process successfully validates the proposed prescriptive approaches.

n MA66 West Bldg 105A ROI of AI: Value and Operational Implications of Artificial Intelligence Sponsored: Artificial Intelligence Sponsored Session Chair: Nigel P. Melville, University of Michigan, Ann Arbor, MI, 48109- 1234, United States 1 - Scientometric Review of Business and Operations AI Literature Nigel P. Melville, University of Michigan, Stephen M. Ross School of Business, 701 Tappan Street, Ann Arbor, MI, 48109-1234, United States Artificial intelligence is rapidly diffusing across industries. Though firms are employing AI in a variety of contexts for a variety of benefits, research examining the business impacts and implications of AI is not abundant. This paper employs scientometrics to identify and synthesize such research and pose a set of research questions to lay the groundwork for future studies of the impacts and implications of AI in business contexts. 2 - Artificial Intelligence: How Customer Reactions Impact Innovation Lisa Yeo, University of California, Merced, Merced, CA, United States, Erik Rolland, Jacquelyn Rees Ulmer, Raymond Patterson Artificial Intelligence (AI) technologies are often included as product features (e.g, facial or voice recognition, autonomous driving systems) that drive product and service innovation. However, such innovations increase software complexity, leading to security and privacy issues. Customer reactions to security or privacy failures may affect product demand; customer demand reaction to the security or privacy implications of new features, such as AI-driven technology, plays a role in regulating the rate of innovation. This work examines the trade-off between product innovation and the increased risk of security breaches in AI-enabled products and services. 3 - Who’s Watching TV Jessica Clark, University of Maryland, Jean-Francois Paiement, Foster Provost This work introduces the novel problem of “user disambiguationö: estimating the likelihood of each member of a small group using a shared account or device. Specifically, we focus on television Set-Top Box viewership data in multi-person households. Our contributions include formulating a predictive problem; developing a method for estimating individuals’ viewership which learns priors for viewership in single-person households then adapts them to the specifics of each multi-person household’s viewership history; and evaluating our new method via a comparison to two ad-hoc heuristics currently used for estimating TV audience composition. 4 - To Predict or Not to Predict: The Case of Inpatient Admissions from the Emergency Department Sriram Somanchi, University of Notre Dame, 344 Mendoza College of Business, University of Notre Dame, Notre Dame, IN, 46556, United States, Idris Adjerid Novel statistical methods have the potential to dramatically improve the efficiency and quality of healthcare. However, dynamic processes and complex infrastructures result in the data with highly variable acquisition costs. We offer a novel and generalizable method that helps reduce the cost while maintaining high accuracy. We utilize this approach in the Emergency Department (ED) context. Specifically, we focus on accurately predicting whether an ED patient will be admitted to an interior hospital unit or discharged from the ED, while minimizing the time-cost of prediction. We show this reduction in time-cost of prediction could translate into a 20% reduction in total time in the ED per patient. n MA67 West Bldg 105B Field Experiment in Online Marketplace and Social Network Sponsored: Information Systems Sponsored Session Chair: Tianshu Sun, University of Southern California 1 - Word-of-mouth System Implementation and Customer Conversion: A Randomized Field Experiment Ni Huang, Arizona State University, 832 W. Wagner Dr, Arizona, Gilbert, AZ, 85233, United States, Tianshu Sun, Pei-yu Chen, Joseph Golden E-commerce firms often face the decision on whether they should implement a word-of-mouth (WOM) system on their websites. An in-site WOM system can potentially boost customer conversion by conveying signals and information

143