Knowledge that Transforms

As the use of information as a productive and tradeable asset becomes more pervasive, its—often unintended—side-effects start showing. This prompts us to systematically recognize these effects and,...

Digital platforms have replaced traditional markets in most industries and orchestrate socioeconomic aspects of our lives. We address the problem of negative direct side network effects that arise with an increased number of agents on one side of the platform. Negative effects, if unaddressed, lead to undesired long-term consequences for the platform by developing a positive vicious cycle. Addressing negative effects require dynamic solution mechanisms that adapt to the changing landscape of platforms. The recommender systems literature has proposed multi-sided recommender systems (MSR) as a dynamic solution to many problems on platforms. However, current state-of-the-art MSRs do not consider uncertainty in predicting agents' choices, resulting in limited efficacy. We present a robust multi-sided recommender system that considers estimation errors in agents' choice to address this concern. Extensive experiments with agent-based models—ride-pooling and education platform—provide support for the efficacy and generalizability of the robust MSR to address negative effects.KEYWORDS: Digital platformsnetwork effectsnegative side effectsmulti-sided platformsmulti-sided recommendersrobust optimizationagent-based simulation AcknowledgmentsWe thank the Editor-in-Chief, Dr. Vladimir Zwass, and the three anonymous reviewers for their constructive suggestions throughout the review process. We also thank participants of the 2019 Winter Conference on Business Analytics (WCBA) and seminar participants at the University of Wisconsin at Milwaukee and Northern Illinois University for their valuable feedback on earlier versions of this paper. Onkar Malgonde acknowledges financial support for this research from the G. Brint Ryan College of Business.Supplementary materialSupplemental data for this article can be accessed online at https://doi.org/10.1080/07421222.2022.2127440Disclosure StatementNo potential conflict of interest was reported by the authors.Notes1 A positive direct side effect refers to the "positive benefits received by users when the number of users of the same kind increases—for example, the effect that arose as the number of subscribers to the Bell Telephone network grew" [40, p. 29].2 One-sided recommender systems have relied on data mining and optimization approaches [Citation41]. For example, Adomavicius and Kwon [Citation1] develop a candidate optimization model to balance diversity with the traditional measure of recommender quality, such as accuracy. As a relatively newer field of research within the recommender systems domain, the MSR literature has extensively relied on optimization models.3 For brevity, this model is based on the method proposed by [Citation45]. In our empirical study, we use the method proposed by Bertsimas and Sim [Citation10] for its efficacy in balancing the optimality of the solution and its protection against constraint violation.4 As a first step in addressing the challenges of uncertainty in data, Soyster [Citation45] proposed a method that traded the optimality of the solution in favor of feasibility for all data. To address this limitation, Ben-Tal and Nemirovski [Citation8] and El-Ghaoui et al. [Citation21] proposed methods that consider robust counterparts of the nominal problem. However, the proposed methods are computationally expensive. To address these computational challenges and retain the optimality of the solution, Bertsimas and Sim [Citation10] propose a method that allows the user to vary the conservatism of the solution. Robust optimization is applied in various fields such as finance (portfolio optimization), supply chain management (inventory control), and engineering design problems [Citation9]. To the best of our knowledge, robust optimization has not yet been introduced to the recommender systems domain.5 Popular platforms include Waze Carpool in the US, BlaBlaCar in Europe, Grab in Southeast Asia, Hitch in China, and Jrney in Africa, among others.6 Similar mechanisms and dynamics of (a) making an offer, (b) accepting/rejecting an offer, (c) using two-way ratings (buyers and sellers rate each other), and (d) agents' objectives, preferences, and constraints are at play on multiple other types of platforms such as lodging marketplaces (Airbnb), freelancing platforms (Upwork, Fiverr), and crowdsourcing platforms (Amazon Turk, Toloka), among others.7 Among other differences, one of the key differences between ride-pooling and ride-hailing platforms is how matches are made: identified by agents (ride-pooling) or determined by the platform (ride-hailing).8 Fitness has been used in prior agent-based simulation studies in the Information Systems domain [Citation20, Citation35, Citation38].9 We thank an anonymous reviewer for recommending this point.10 https://www.forbes.com/sites/jeffbercovici/2014/08/14/what-are-we-actually-rating-when-we-rate-otherpeople/?sh=78550e4debca11 https://www.businessinsider.com/leaked-charts-show-how-ubers-driver-rating-system-works-2015-212 https://www.uber.com/en-TW/blog/5starriders/13 https://www.uber.com/newsroom/rider-ratings-breakdown/14 An earlier version of the manuscript used flexible tuples (e.g. [Citation1,0,0,Citation1,0]) to represent a focal agent's preferences. To generate the preference scores between two agents, we took the ratio of the number of common elements over the tuples' length (e.g., for a rider [Citation1,0,0,Citation1,0] and driver [Citation1,Citation1,Citation1,0,0], the preference match score is 2/5 = 0.4). Although we note the heterogeneity of ratings for a focal driver across all riders using this approach, the mean rating for all drivers is within the range [0.48, 0.53] and may not represent the preference dynamics on ride-pooling platforms. We appreciate the comments of an anonymous reviewer in identifying this point.15 We choose 0.98 as the default parameter because it translates to an average rating of 4.9 (the average rating of riders and drivers based on anecdotal evidence) on a 5-point scale used by ride-pooling platforms.16 A random value drawn from a uniform distribution; redrawn for each composite measure in a period; the same for all drivers/riders across all simulation universes for a period.17 A random fraction of the existing fitness value; the same for all drivers/riders across all simulation universes.18 For example, we assume that a focal driver's fitness at the start of a period is 0.6. Also, assume that the average capacity utilization across all rides offered by the focal driver exceeds a specified valuexv, and that the focal driver's average quality of a match is greater than a specified valuexv. If we determine that the fitness increment should be 0.25xvii, then the fitness of the focal driver will be (0.6+(0.6*0.25)) = 0.75. We use a similar logic for riders.19 Although unavailable for ride-pooling platforms, estimates from a related platform suggest a ratio of 20 riders per driver (https://www.earnestresearch.com/behind-the-ridesharing-wheel/; accessed: 05/10/2022)20 In Tables 3 and 4, Gk is the average fitness of agents in the k universe, where k is either no recommender, a one-sided recommender, a multi-sided recommender, or a robust multi-sided recommender system. Each cell is a Wilcoxon signed-rank test comparing agents' fitness between the Robust Multi-sided Recommender System and the other system. When statistically significant, we conclude that the difference between the two systems' fitness is statistically significant and that the system with greater average fitness (Gk) outperforms the other system. The Wilcoxon signed-rank test is appropriate because (a) we need to compare systems' performance across multiple simulation runs and (b) acomparison should be paired to a simulation run—initial parameters of a simulation running across different systems are the same.21 We compute Cohen's d for each pair of compared systems. The literature suggests that the effect size can be small (d ≤ 0.2), medium (d = 0.5), or large (d ≥ 0.8), and provides broad categories that should be informed by the study's context [Citation46]. Each cell shows the probability of superiority and Cohen's d for the compared systems.22 We thank Dr. Vladimir Zwass, JMIS Editor-in-Chief, and the three anonymous reviewers for this discussion.Additional informationNotes on contributorsOnkar S. MalgondeOnkar S. Malgonde ([email protected]; corresponding author) is an Assistant Professor in the Information Technology & Decision Sciences Department, G. Brint Ryan College of Business, University of North Texas. He received his Ph.D. in Information Systems from the University of South Florida. Before starting his graduate studies, Dr. Malgonde was a Systems Engineer with Infosys Technologies. His research interests are at the intersection of data analytics, digital platforms, and software systems. His work has been published in such journals as MIS Quarterly, Empirical Software Engineering, and Electronic Markets, and in the proceedings of premier Information Systems conferences and workshops.He ZhangHe Zhang ([email protected]) is an Assistant Professor in the Information Systems and Decision Sciences Department in the Muma College of Business at University of South Florida. His research interests include healthcare information management, big data, and production and inventory management. Dr. Zhang's research has been published in several journals, including MIS Quarterly, Mathematical Programming, Decision Support Systems and ACM Transactions on Management Information Systems.Balaji PadmanabhanBalaji Padmanabhan ([email protected]) is the Anderson Professor of Global Management in the Information Systems Decision Sciences Department and Director of the Center for Analytics & Creativity at University of South Florida. He received his Ph.D. from the Stern School of Business of New York University. Dr. Padmanabhan's interests include analytics and business intelligence, designing analytics algorithms for business applications, building and evaluating predictive models, patterns discovery in data, enabling citizen data science and applications of analytics in healthcare, recommender systems, fraud detection and elections. His research has been published in the premier Computer Science and Business journals and conference proceedings. He serves on the editorial boards and program committees of many leading academic journals and conferences.Moez LimayemMoez Limayem ([email protected]) is the President of the University of North Florida. Until June 2022, he was the Lynn Pippenger Dean of the Muma College of Business at the University of South Florida, which he joined coming from the Sam M. Walton College of Business at the University of Arkansas. Dr. Limayem's research was published in many top journals, such as Information Systems Research, Journal of Management Information Systems, MIS Quarterly, Journal of the AIS, and Management Science.

Organizations usually have difficulty adjusting to technology-enabled changes. Recent research has examined the interaction between technology and the key job outcomes of employees. But this research stream has done so using a linear lens even though this interplay has been recognized to be dynamic and complex. We challenge here this linearity assumption. We theorized that enterprise system (ES) use influences post-implementation job scope, and the change from pre- to post-implementation job scope perceptions will have a complex effect on job outcomes that are best captured by a polynomial model. Drawing on the anchoring-and-adjustment perspective in decision-making research, our polynomial model highlights the dynamic nature of employee reactions to changes in job scope brought about by an ES implementation that cannot be captured by traditional linear models. We found support for our model using data collected in a longitudinal field study from 2,794 employees at a telecommunications firm over a period of 12 months. Our findings highlight the key role an ES implementation can have in changing the nature of jobs and how those changes can, in turn, drive job performance and job satisfaction. This research also extends classical job characteristics research by arguing for a more complex relationship between the scope and outcomes of technology-supported jobs.

Worker persistence—the ability to focus on a task for long periods of time—is often highlighted as essential to success. However, computers are extraordinarily persistent, particularly for routine, repetitive work. This potentially reduces the value of human persistence in occupations that are computerized. Using a well-defined measure of worker persistence across a nationally-representative 16-year sample of 4,239 individuals, we investigate the extent to which occupations value worker persistence in the presence of computers. We find that the labor market does indeed value persistence. Nonetheless, we find that in routine jobs, the wage premium of human persistence diminishes with the degree of workplace computerization. Yet, this substitution does not occur in non-routine jobs. These findings deepen our understanding of the effect of workplace computerization on the future of work and workers, and they also warrant imlications on government job training programs, organizational talent management, as well as the redesign of the K-12 curriculum.

Reward-based crowdfunding has become a popular fundraising marketspace for entrepreneurs. However, few studies take a process perspective and examine how fundraiser network influences the execution effectiveness of crowdfunding projects. Collecting the funding activities of all fundraisers on Kickstarter in 2017, we construct a fundraiser network and examine the effects of fundraiser network structural characteristics on project execution effectiveness. Employing endogenous stochastic frontier analysis, we show that reciprocity and accessibility in the fundraiser network increase project execution effectiveness. We also demonstrate that gaining network status by virtue of being an authoritative figure in the fundraiser network increases project execution effectiveness. In sum, our study identifies the effects of the fundraiser network characteristics on project execution effectiveness and introduces a process perspective for evaluating crowdfunding campaigns.

Machine learning is widely used in information systems design. Yet, training algorithms on imbalanced datasets may severely affect performance on unseen data. For example, in some cases in healthcare, fintech, or cybersecurity contexts, certain subclasses are difficult to learn because they are underrepresented in training data. Our study offers a flexible and efficient solution based on a new synthetic average neighborhood sampling algorithm (SANSA), which, in contrast to other solutions, introduces a novel "placement" parameter that can be tuned to adapt to each dataset's unique manifestation of the imbalance. This package can be downloaded for RFootnote1. We tested SANSA against seven existing sampling methods used in conjunction with the four most frequently used machine learning models trained on 14 benchmark datasets. Our results provide suggestive evidence that SANSA offers a feasible solution to the imbalance problem for most datasets. Our findings provide practical recommendations for how SANSA can be effectively implemented while reducing the complexity level of an imbalanced learning pipeline.KEYWORDS: Imbalanced dataoversamplingundersamplingmachine learningpredictive analyticsclassification prediction performancealgorithm training AcknowledgmentEarlier versions of this research work that introduced our novel SANSA algorithm for the first time had received various awards such as the Winner of the international-level Decision Sciences Institute (DSI) Regional Best Paper Award at the DSI 2020 Annual Conference across all six chapters of DSI globally, the Finalist for the Best Paper Award in Data Mining at the INFORMS 2020 Annual Conference, Best Contribution to Theory Paper Award at the NEDSI 2020 Annual Conference, and the Best Overall Conference Paper Award at the NEDSI 2020 Annual Conference. We are grateful to all judges and award committee members for their evaluation and consideration of our work worthy of such awards. We are indebted to the audiences of those conferences for their invaluable feedback to further improve this work. We are also thankful to the three anonymous referees of this JMIS review panel for their comments and suggestions that significantly helped us further improve our work. Last, but not the least, we thank Editor-in-Chief Dr. Vladimir Zwass for his very timely management of our submission and his constructive comments throughout the submission process.Supplementary informationSupplemental data for this article can be accessed online at https://doi.org/10.1080/07421222.2022.2127453Disclosure StatementNo potential conflict of interest was reported by the authors.Notes1. https://cran.r-project.org/web/packages/sansa/2. Following the suit of other studies, we use Euclidian space to calculate distances, which, if needed, can be replaced with other distance metrics using different geometry within SANSA or any of the other ML algorithms.3. https://murtaza.cc/SANSA/#f5d4. https://murtaza.cc/SANSA/#f5e5. Anirban Datta, Personal Loan Modeling - https://www.kaggle.com/teertha/personal-loan-modelingAdditional informationNotes on contributorsMurtaza NasirMurtaza Nasir ([email protected]) is a Ph.D. candidate in Management Science at the Manning School of Business, University of Massachusetts Lowell and an incoming Assistant Professor at the Finance, Real Estate, and Decision Sciences Department at the W. Frank Barton School of Business, Wichita State University. His research interests are in machine learning and data mining, and span both theory as well as application. He has worked on healthcare, finance, and operational predictive and prescriptive analytics, in addition to pedagogical and theoretical work in business analytics and machine learning. He has received Best Paper awards at DSI 2020, NEDSI 2020, and NEDSI 2021 and was a finalist at the INFORMS 2020.Ali DagAli Dag ([email protected]) is an Associate Professor of Analytics at the Business Intelligence & Analytics Department at the Heider College of Business, Creighton University. He received his Ph.D. from Auburn University. His research interests include business and data analytics, operations research, operations management, and text mining. Dr. Dag is serving as an associate editor of Journal of Business Analytics, Journal of Modeling in Management, and AI in Business journals. His research work has been published in many journals, such as Journal of Management Information Systems, Decision Support Systems, OMEGA: The International Journal of Management Science, Annals of Operations Research, Journal of Business Research, Information Systems Frontiers, among others. He has received Best Paper awards at DSI 2020, NEDSI 2020, and was a finalist at the INFORMS 2020 conferences.Serhat SimsekSerhat Simsek ([email protected]) is an Assistant Professor in the Department of Information Management & Business Analytics at the Feliciano School of Business, Montclair State University. He earned his Ph.D. in Statistics from Auburn University. His research interests span information systems, healthcare analytics, and machine learning. His work has appeared in such journals as Journal of Management Information Systems, OMEGA, Decision Support Systems, Annals of Operations Research and several others. His research has received awards from leading conferences including INFORMS and DSI.Anton IvanovAnton Ivanov ([email protected]) is an Assistant Professor of Information Systems in the Gies College of Business, University of Illinois at Urbana-Champaign. He received his Ph.D. in Management Information Systems from the State University of New York at Buffalo. Dr. Ivanov's research stands at the intersection of information systems, social media, and healthcare analytics with an emphasis on the user-generated content.Asil OztekinAsil Oztekin ([email protected], *corresponding author) is an Associate Professor of Analytics & Operations Management in Manning School of Business at the University of Massachusetts Lowell. He earned his Ph.D. from Oklahoma State University. Dr. Oztekin's research interests relate to data science, data mining, predictive analytics, decision analytics, decision support systems with applications in healthcare analytics, marketing analytics, and text mining. He has published over 50 peer-reviewed articles in the leading journals and conference proceedings, including Journal of Management Information Systems, European Journal of Operational Research, Decision Support Systems, International Journal of Production Research, OMEGA, Information Systems Frontiers, and Annals of Operations Research, among others. He serves as senior editor/associate editor/ editorial review board member for Journal of the Association for Information Systems, Decision Sciences, European Journal of Operational Research, Decision Support Systems, Journal of Business Research, and others. His research work has received several awards from various venues, such as DSI Annual Conference, Northeast Decision Sciences Institute, and INFORMS.

Peer-to-peer (P2P) rental markets have been shown to adversely impact the traditional hospitality industry and housing affordability, fueling the public demand for regulation. While localities around the globe have implemented policies to address these issues, little is known about how rental suppliers respond to those regulations. This study aims to empirically analyze the impact of such policy regulations on the prices charged by different types of rental suppliers. We employed a quasi-experimental research design based on an extensive dataset including more than 50,000 Airbnb listings to uncover the impact of a policy implemented in New Orleans, which introduced annual bring-to-market (BTM) costs through a mandatory licensing system while simultaneously banning listings from one city-center neighborhood. We find that, while non-commercial hosts completely pass their additional costs onto their consumers, irrespectively of demand and supply shifts, commercial hosts’ responses are more nuanced. Those with legalized listings located in the city center only partially pass on their costs to guests, while even decreasing their prices in the rest of the city. Our study contributes to the understanding of pricing in P2P rental markets and its effects. Further, it informs localities and supports policy analytics. With P2P renting remaining attractive in city parts where BTM costs can easily be passed through to consumers, this suggests that these regulatory policies fall short of reducing pressure on housing affordability in the city-center.

We aim to understand why employees take information technology (IT)-mediated shortcuts, that is, skipping one or several steps for completing tasks quicker by bending the rules. This is a specific and often detrimental form of noncompliant behavior. Adopting an ego-depletion perspective, we posit that IT complexity drives IT-mediated shortcuts by increasing employees’ ego-depletion. Extending this view, we use a modified Delphi study and build on self-regulatory and goal setting theories to point to key boundary conditions for these effects. First, in a preliminary study we found that taking IT-mediated shortcuts in our context is, on average, detrimental to employee performance. This highlighted the need to focus on IT-mediated shortcuts. Next, we tested our assertions with three experiments focusing on the use of dashboards with 584 data analysts. The results show that (1) dashboard complexity increases ego depletion, (2) ego depletion fully mediates the impact of dashboard complexity on taking IT-mediated shortcuts, (3) moral integrity moderates the influence of ego depletion on taking IT-mediated shortcuts, and (4) outcome compared to learning goals enhance the impact of ego depletion on IT-mediated shortcuts. In all studies, objectively measured IT-mediated shortcut-taking was negatively associated with objectively measured task performance. Ultimately, the integrated perspective explains whether, how, and under what conditions IT complexity drives IT-mediated shortcuts.

Some people contribute ideas for prosocial reasons in crowdsourcing; others do so for selfish reasons. Extending the theory of motivated information processing, the research posits that prosocial and proself individuals respond differently to reward structures in crowd idea generation. Two online experiments measured participants' prosocial versus proself orientation and manipulated whether participants received a competitive or cooperative reward structure. Study 2 also manipulated whether participants viewed an original or a common peer idea. Proselfs produced more ideas when receiving competitive rewards; the idea generation of prosocials was not affected by the reward structure. This interaction effect was mediated by task effort and moderated the impact of peer ideas. Proselfs generated the most ideas when viewing an original peer idea and receiving competitive rewards; this effect was not observed for prosocials. The study contributes to crowdsourcing research by demonstrating that participants' response to reward structures depends on their social value orientation. The implication is that crowdsourcing organizers should design tasks and rewards so they motivate participants with both prosocial and proself orientations.

Doctors and paraprofessionals operate in stressful environments that jeopardize their well-being and quality of care. E-healthcare systems have been promoted by government initiatives (e.g., HITECH act) to support healthcare services. Recent evidence suggests, however, that these systems contribute to job strain. Drawing on findings from a qualitative study and proximity and homophily theories, we integrate the healthcare context to develop and test a research model of friendship network ties among and between doctors and paraprofessionals as a coping mechanism for alleviating job strain. We test our model in a year-long field study in a hospital that implemented a new e-healthcare system, with two waves of data collected from 152 doctors and 731 paraprofessionals. Our findings move beyond the classical view of friendship as a conduit of coping by suggesting that the source of friendship network ties could reduce or aggravate strain as doctors and paraprofessionals continue to interact with the system over time.KEYWORDS: Job strainjob copinge-healthcare usefriendship tieshomophily theoryproximity theory Supplementary materialSupplemental data for this article can be accessed online at https://doi.org/10.1080/07421222.2022.2127450Disclosure StatementNo potential conflict of interest was reported by the authors.Notes1. Social network data used for the analysis of this period were collected at the end of the fifth month. During the second month to the fifth month, the number of employees' friendship ties was expected to change because the new system could trigger interactions that could create new connections.2. We estimated the effects for the periods T5 and T12 and found them to be consistent. Hence, the results were reported based on data pooled across these time periods.Additional informationNotes on contributorsTracy Ann SykesTracy Ann Sykes ([email protected]) is an associate professor of Information Systems in the Sam M. Walton School of Business at the University of Arkansas. She has been a member of the faculty of the Research School of Business in the College of Business and Economics at The Australian National University and has worked as a science assistant at the National Science Foundation. Dr. Sykes's research focuses on leveraging social network theory, methods, and analyses to understand technology-related phenomena. She also works in the contextual areas of organizational and societal diffusion of technologies in developing countries and in health IT. Her work has been published in leading journals, such as MIS Quarterly, Information Systems Research, Academy of Management Journal, Journal of the American Medical Informatics Association, Production and Operations Management, and Journal of Applied Psychology.Ruba AljafariRuba Aljafari ([email protected]; corresponding author) is an assistant professor of Business Information Technology in the Pamplin College of Business at Virginia Tech. She completed her Ph.D. in Information Systems at the Sam M. Walton College of Business, University of Arkansas. Dr. Aljafari's research interests include ICT implementation in organizations, usability, and healthcare IT, with an emphasis on patient-centered e-health and analytics. Her work has been published in leading journals in information systems, healthcare, and human-computer interaction, such as MIS Quarterly, Journal of the American Medical Informatics Association, and the International Journal of Human-Computer Studies.