Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

SB52

Panelists Renata Alexandra Konrad, Worcester Polytechnic Institute, School of Business, 100 Institute Road, Worcester, MA, 01609, United States Paul R. Messinger, University of Alberta, Faculty of Business, 3- 20e Faculty Of Business Bldg, Edmonton, AB, T6G 2R6, Canada Richard G. McGrath, United States Naval Academy, 572C

5 - Project Resource Leveling Under Uncertainty Hongbo Li, Shanghai University, Baoshan District, Shangda Road 99, Shangchuan Road 995, Shanghai, 200444, China We investigate project resource leveling with stochastic activity durations as well as resource availabilities to minimize the fluctuations in resource usage. To obtain effective scheduling policies for the stochastic resource leveling problem, we model the problem as a Markov decision process model and propose approximate dynamic programming algorithms built upon Metaheuristics. To evaluate our algorithms, we conduct extensive computational experiments on benchmark instances. n SB52 North Bldg 231C Joint Session Award/Practice Curated: Social Media Analytics Best Student Paper Competition Emerging Topic: Social Media Analytics Emerging Topic Session Chair: Julie Zhang, University of Massachusetts, Lowell, Operations and Information Systems, Lowell, MA, United States 1 - Detecting Influence Campaigns in Social Networks Using the Ising Model Nicolas Guenon des Mesnards, Massachusetts Institute of Technology, Cambridge, MA, USA, . We consider the problem of identifying coordinated influence campaigns conducted by automated agents or bots in a social network. We study several different Twitter datasets which contain such campaigns and find that the bots exhibit heterophony - they interact more with humans than with each other. We use this observation to develop a probability model for the network structure and bot labels based on the Ising model from statistical physics. We present a method to find the maximum likelihood assignment of bot labels by solving a minimum cut problem. Our algorithm allows for the simultaneous detection of multiple bots that are potentially engaging in a coordinated influence campaign, in contrast to other methods that identify bots one at a time. We find that our algorithm is able to more accurately find bots than existing methods when compared to a human labeled ground truth. We also look at the content posted by the bots we identify and find that they seem to have a coordinated agenda. 2 - Detecting Changes in Dynamic Events over Networks Shuang Li, Georgia Institute of Technology, Atlanta, GA, United States Large volume of networked streaming event data are becoming increasingly available in a wide variety of applications, such as social network analysis, Internet traffic monitoring and healthcare analytics. How to promptly detect changes in these dynamic systems using these streaming event data? In this paper, we propose a novel change-point detection framework for multi- dimensional event data over networks and cast the problem into sequential hypothesis test. We show that our method can achieve weak signal detection by aggregating local statistics over time and networks. Finally, we demonstrate the good performance of our algorithm on numerical examples and real-world datasets from twitter and Memetracker. 3 - How do Sales Responses to Various User-generated-content? A Panel VAR Analysis based on Twitch, YouTube and Steam Data Shuang Li, Georgia Institute of Technology, Atlanta, GA, United States Twitch.tv (Amazon owned), having over 15 million daily active users, owning over 43 percent of game video content market and ranking 4th in in peak US internet traffic, is the biggest live stream platform now. The unique features of live stream video platforms include timely broadcasting and interpersonal interaction, large scale user base and democratized content genre distribution. Given these unique features, Twitch has been more and more popular among various marketers across different industries as a new form of marketing channel which have timely impacts and high efficiency. Meanwhile, pre-recorded video platforms like YouTube is also largely used as an effective marketing channel. Thus, it is unclear that which platform as better marketing efficiency. Additionally, compared with traditional user-generated-content (UGC) like reviews, it is unclear that whether user-generated video contents about a product on two platforms are associated with the product’s reputation in terms of reviews- the traditional form of user generated content. To explore these two concerns, we merge a unique dataset including game sales and profile data from the largest online game distribution platform- Steam, game related data from YouTube and Twitch and conduct a comprehensive Panel VAR analysis. We provide both empirical and theoretical implications based on summarized finding patterns and illustrate contributions of this study.

Holloway Rd, Annapolis, MD, 21402, United States Eric Johnson, Vanderbilt, Nashville, TN, United States

n SB51 North Bldg 231B Advances in Project Management and Scheduling Emerging Topic: Project Management and Scheduling, in Memory of Joe Leung, Emerging Topic Session Chair: George Vairaktarakis, Case Western Reserve University, Cleveland, OH, 44106-7235, United States 1 - The Inventory vs. Timeliness Tradeoff in Project Delivery Arman Jabbari, University of California-Berkeley, 2012 Del Norte, Berkeley, CA, 94707, United States, Phil Kaminsky We explore the tradeoffs between inventory cost and project completion times in a variety of settings, across single and multiple projects. 2 - Managing Clinical Trials in a Drug Development Project Theodore D. Klastorin, University of Washington, ISOM Department, Box 353226, Seattle, WA, 98195-3226, United States, Kamran Moinzadeh, Hamed Mamani We consider the case when a series of clinical trials are needed to validate a new drug under development. The issue we consider is related to the number of patients to enroll in each trial and when these patients should be enrolled. We show how this problem is related to flexible resource allocation problems in project management and develop a model that analyzes static versus dynamic scheduling strategies. 3 - Information Asymmetry in Budget Allocation: An Analysis of a Truth-inducing Incentive Scheme Yun Zhou, 1989, NDSU, Fargo, ND, United States, Joseph Szmerekovsky Truth-inducing incentive schemes are used to motivate project managers to provide unbiased project information to the portfolio manager to reduce information asymmetry between the portfolio manager and the project managers. To improve the scheme, we identify the proper value of penalty coefficients in the truth-inducing incentive scheme when information asymmetry is present. We first describe the allocation method that achieves budget optimization under certain assumptions and identify the proper coefficients while accounting for the differing perceptions of both the portfolio manager and the project managers. We report a bound on the ratio between the two penalty coefficients in the truth- inducing incentive scheme. We conclude that the penalty coefficient for being over budget should be reduced when the portfolio budget is tight and the penalty coefficients should be equivalent to the organizational opportunity costs when the portfolio budget is sufficient. 4 - A Cutting Plane Approach for the Multi-Machine Precedence- Constrained Scheduling Problem George Vairaktarakis, Case Western Reserve University, Dept of OR and OM, 10900 Euclid Avenue, Cleveland, OH, 44106-7235, United States, Prahalad Venkateshan, Joseph Szmerekovsky A cutting-plane approach is developed for the problem of optimally scheduling jobs with arbitrary precedence constraints on unrelated parallel machines. Our model utilizes a number of valid inequalities that cut off fractional linear programming solutions. This leads to an increase of the linear programming lower bound from 89.3% to 94.6% of the corresponding optimal solution, and a substantial reduction in the computational time of an optimal branch-and-bound algorithm for this problem. We report optimal solutions for problem instances with up to 25 jobs and 5 machines; more than twice the size of problems for which optimal solutions have been reported so far.

49