2016 INFORMS Annual Meeting Program

TB66

INFORMS Nashville – 2016

2 - Free Riders Versus Social Capital: An Empirical Analysis Of An Exogenous Shock On Online Reviews Zaiyan Wei, Purdue University, West Lafayette, IN, United States, zaiyan@purdue.edu, Paulo B Goes, Yang Wang, Dajun Daniel Zeng We study the effects of network sizes on individuals’ contributions to online product reviews. Individuals have conflicting incentives of free riding and maximizing social benefits when producing online reviews. We leverage a “natural experiment,” an exogenous expansion in the users population on a major third-party platform, to better understand the tradeoffs between the conflicting incentives. We find that a larger population of users caused individuals to post more and longer reviews. In addition, the larger population of audience led individuals to assign higher and more diverse ratings in their reviews. However, the helpfulness or “quality” of reviews is not affected. 3 - Enterprise Systems And Merger And Acquisition Activities Chengxin Cao, University of Minnesota, 321 Nineteenth Avenue South, Minneapolis, MN, United States, caoxx161@umn.edu, Gautam Ray, Alok Gupta, Mani Subramani This paper examines the relationship between Enterprise Resource Planning (ERP) and Customer Relationship Management (CRM) systems and upstream and downstream mergers and acquisitions (M&A). We also investigate how any such relationship is contingent on the characteristics of the focal firms’ industry environment. Using a sample of 491 Fortune 1000 firms that made 4543 M&A deals from 2006 to 2012 the empirical analysis suggests that ERP (CRM) systems are negatively associated with upstream (downstream) M&A. However, if the upstream (downstream) industry is concentrated (dynamic), ERP (CRM) systems are associated with more vertical M&A. 4 - The Influences And Biases Of Social Network In Referral Hiring: Empirical Study Kyungsun Rhee, University of Washington, 4725 24th Avenue NE, # 405, Seattle, WA, 98105, United States, ksr22@uw.edu, Elina Hwang, Param Vir Singh It is well known that importance of social networks in labor market has been growing rapidly. However, there have been rigorous researches on characteristics of job seekers who are likely to achieve better results in job market, but not many on the referrer behavior. Using data from social referral platform, this paper constructs an empirical model to capture the influences and biases of referrers’ social capital on their actual referring behavior in the IT labor market. TB66 Mockingbird 2- Omni Data Analytics for Quality and Reliability Assurance Sponsored: Quality, Statistics and Reliability Sponsored Session Chair: Mingyang Li, Tampa, FL, United States, mingyangli@usf.edu 1 - A Data-driven Heterogeneity Quantification Approach For Chloride Ingress Profiles Of Aging Marine Infrastructures Suiyao Chen, University of South Florida, 4202 E. Fowler Avenue ENB302, Tampa, FL, 33620, United States, suiyaochen@mail.usf.edu, Lu Lu, Yisha Xiang, Alberto A Sagüés, Mingyang Li Chloride ingress is the leading cause to corrosion failures of aging infrastructures in marine environments. Existing studies on chloride ingress mainly assumed homogeneous populations and were constrained by the simplified physical assumptions and availability of chloride ingress profiles. In this work, a data- driven approach is presented to comprehensively explore, quantify and analyze the heterogeneous chloride ingress profiles collected from a field survey on marine infrastructures. A real-world case study is provided to illustrate the proposed work and demonstrates its validity and performance. 2 - Reliability Meets Big Data: Opportunities And Challenges Yili Hong, Virginia Tech, yilihong@vt.edu In this talk, I will review some applications where field reliability data are used and explore some of the opportunities to use modern reliability data to provide stronger statistical methods to operate and predict the performance of systems in the field. I will also provide some examples of recent technical developments designed to be used in such applications and outline remaining challenges.

3 - Heterogeneous Recurrence Representation And Quantification Of Dynamic Transitions In Continuous Nonlinear Processes Hui Yang, Penn State, huy25@engr.psu.edu Many real-world systems are evolving over time and exhibit dynamical behaviors. In order to cope with system complexity, sensing devices are commonly deployed to monitor system dynamics. Online sensing brings the proliferation of big data that are nonlinear and nonstationary. Although there is rich information on nonlinear dynamics, significant challenges remain in realizing the full potential of sensing data for system control. This paper presents a new approach of heterogeneous recurrence analysis for online monitoring and anomaly detection in nonlinear dynamic processes. 4 - Latent Dirichlet Allocation (lda) Based Analytic Framework For Topic Modeling Of Cfpb Consumer Complaints Kaveh Bastani, Recovery Decision Science, Cincinnati, OH, United States, kaveh@vt.edu, Hamed Namavari, Jeffrey Shaffer We propose a text mining analytic framework based on latent Dirichlet allocation (LDA) to analyze Consumer Financial Protection Bureau (CFPB) consumer complaints. The proposed analytic framework aims to extract latent topics/clusters in CFPB complaint narratives, and explores their associated trends over time. The time trends will then be used to evaluate the quality of industry regulations and expectations on financial institutions in creating a consumer oriented culture that takes into account consumer protection in their decision making processes. IIE Transactions Invited Session Sponsored: Quality, Statistics and Reliability Sponsored Session Chair: Jianjun Shi, Georgia Institute of Technology, Atlanta, GA, United States, jianjun.shi@isye.gatech.edu 1 - A Random Effect Autologistic Regression Model With Application To The Characterization Of Multiple Microstructure Samples Qingyu Yang, Wayne State University, qyang@wayne.edu The microstructure of the material can strongly affect material properties which in turn plays an important role of the product quality produced by these materials. The existingstudies on material microstructure mainly focus on a single microstructure sample’s characteristics, while the variation among different microstructure samples is ignored. In this paper, we propose a novel random effect autologistic regression model to characterize the microstructure variation of different samples for the two phase materials. A simulation study is conducted to verify the proposed methodology. A real world example of a dual-phase high strength steel is used to illustrate the developed methods. 2 - A Bayesian Variable Selection Method For Joint Diagnosis Of Manufacturing Process And Sensor Faults Yong Chen, University of Iowa, Iowa City, IA, 52242, United States, yong-chen@uiowa.edu This paper presents a Bayesian variable selection based diagnosis approach to identify both process mean shift faults and sensor mean shift faults simultaneously in manufacturing processes. Important concepts are introduced to understand the diagnosability of the proposed method. A conditional maximum likelihood method is proposed as an alternative method to provide robustness to selection of some key model parameters. Systematic simulation studies are used to provide insights on the relation between the success of the diagnosis method and related system structure characteristics. And a real assembly example is used to demonstrate the effectiveness of the proposed diagnosis method. 3 - A Preposterior Analysis To Predict Identifiability In The Experimental Calibration Of Computer Models Daniel Apley, Northwestern University, apley@northwestern.edu When calibrating computer simulation models using physical experimental data, it is usually very difficult to identify unknown physical parameters and distinguish their effects from the discrepancy function that represents the difference between the simulation model and reality. We develop a preposterior analysis to predict (prior to conducting physical experiments but after conducting simulations) the identifiability that will result for any candidate physical experimental design. This can be used as a criterion for designing physical experiments to achieve better identifiability of the physical calibration parameters. TB67 Mockingbird 3- Omni

286