Jeffrey Parsons, University Research Professor and Professor of Information Systems, Faculty of Business Administration, Memorial University of Newfoundland

Personalizing Online Reviews for Better Customer Decision Making. Online consumer reviews have become an important source of information for understanding markets and customer preferences. When making purchase decisions, customers increasingly rely on user-generated online reviews, which in some cases are more credible and trustworthy than information provided by vendors. Many studies have revealed that online reviews influence demand and sales. Others have shown the possibility of identifying customer interest in product attributes. However, little work has been done to address customer and review diversity in the process of examining reviews. This research intends to answer the research question: how can we solve the problem of customer and review diversity in the context of online reviews to improve product recommendation based on customer preferences? Our approach to the question is through personalization. Similar to other personalization research, we use an attribute-based model to represent products and customer preferences. Unlike existing personalization research that uses a set of pre-defined product attributes, we use a data-driven approach to identify a more comprehensive set of product attributes from online reviews to model products and customer preferences. Specifically, we introduce a new topic model for product attribute identification and sentiment analysis from textual review. By differentiating word co-occurrences at the sentence level from co-occurrences at the document level, the model better identifies interpretable topics. The use of an inference network with shared structure enables the model to predict product attribute ratings accurately. Based on this topic model, we develop attribute-based representations of products, reviews and customer preferences and use them to personalize online reviews. We hypothesize the effects of personalization from the lens of consumer search theory and human information processing theory and test these hypotheses in an experimental setting. The empirical evidence shows that the personalization of online reviews can: 1) recommend products matching customer’s preferences; 2) improve custom’s intention towards recommended products; 3) best distinguish recommended products from products that do not match customer’s preferences; and 4) reduce decision effort.

Brian Butler, Professor and Senior Associate Dean, University of Maryland / College of Information Studies

A Potential Uses Model of Data Governance. As the scale, scope, and pervasiveness of data resources increase, so to do examples of radical new uses and new risks arising from use of data. As a result, researchers and practitioners are increasingly interested in understanding the challenges of data governance. However, much of the published work on data governance is atheoretical, focusing on general advocacy for the importance of data governance; rich description of specific cases; or empirically-derived descriptions of data governance goals. While this literature provides suggestive insights, it is fundamentally limited as a conceptual basis for research and practice. Drawing together a conceptualization of governance as n-steps removed shaping of action, the consequences of the materiality of data, and an assumption of bounded collective rationality, the Potential Uses Model of Data Governance seeks to provide a foundational model of data governance. In this model (and the associated formalism), data governance is described as an effort to indirectly shape the outcomes of data use, promoting those that are desirable and hindering those that are undesirable. Examination of a basic formal model based on the Potential Uses Model of Data Governance highlights fundamental tensions that exist within in the simplest contexts and suggests potential avenues for empirical study, practical analysis, and additional theoretical work.

Gregory Vial, Membre du GReSI et Professeur adjoint, Département des TI, HEC Montréal, & Suzanne Rivard, Membre du GReSI, Professeure titulaire et Titulaire de la Chaire de gestion stratégique des TI, Département des TI, HEC Montréal

Vers une conceptualisation du développement de systèmes comme routine organisationnelle : Implications pour la recherche en TI. La recherche en développement de systèmes d’information (DSI) repose sur une compréhension générale du DSI comme un phénomène incluant à la fois une idée quant à la façon dont il doit être performé, ainsi que la performance du DSI en pratique. Ce projet de recherche a pour but de développer une conceptualisation du DSI reposant sur une ontologie permettant de tenir explicitement compte de cette réalité. À cette fin, nous suggérons une conceptualisation du DSI comme routine organisationnelle basée sur l’ontologie proposée par Feldman et Pentland. Au cœur de cette ontologie, la nature mutuellement constitutive des aspects ostensif et performatif qui composent toute routine organisationnelle nous permet de développer des arguments théoriques sur l’interrelation existant entre (1) le DSI comme idée (l’aspect ostensif), le DSI en pratique (l’aspect performatif), et (3) les artefacts qui sont créés et mobilisés par les acteurs impliqués dans le DSI. Ces arguments servent par la suite de plateforme pour le développement d’un agenda visant à étudier des enjeux importants du DSI contemporain et ayant pour but de favoriser le développement de nouvelles contributions théoriques propres à la recherche en TI pour le développement de systèmes d’information.

Ofir Turel, Professor of Information Systems and Decision Sciences, College of Business and Economics / California State University, Fullerton, and
Scholar in Residence at the Decision Neuroscience Program, Department of Psychology, niversity of Southern California (USC)

Manipulating Online Risk Behaviors through Foraging Theory Insights. Animals change their risk preferences during foraging based on their perceived need for caloric energy intake; they become more risk prone when the need is exacerbated. Humans follow the same pattern concerning financial rewards; when inequality in financial gains is high, people feel that they are dissatisfied with their gains, they need more money to be satisfied, and ultimately take larger financial risks. Given the growing prevalence of online risk taking (e.g., taking selfies on the edge of a cliff, and posting selfies while driving) and the need to understand and prevent such behaviors, I extend the abovementioned ideas rooted in foraging and risk-sensitivity theories to explain online risk taking by information systems (IS) and specifically social media users. I test these assertions in three studies (total of 1,051 social media users), all of which manipulate inequality of information, and each of which focuses on a different “online foraging” behavior (chasing after “likes” through profile picture posting, birthday wishes through birthday picture posting, and social attention through sharing politically-oriented opinion pieces, in studies 1, 2 and 3, respectively). The findings quite consistently support the assertions and provide important theoretical and practical implications.

Yves Pigneur, Professor of Management and Information Systems, HEC Lausanne

A Design Theory for Visual Inquiry Tools. The Business Model Canvas opened the way for the development of a new tool type which we call visual inquiry tools. Such tools build on design thinking techniques to allow management practitioners to jointly inquire into specific strategic management problems. As the interest in and the emergence of visual inquiry tools gains momentum, it is important to formalize the design knowledge that future designers can build on for developing such tools. Thus, we propose a design theory for visual inquiry tools based on the design knowledge accumulated within and across three projects: the Business Model Canvas, the Value Proposition Canvas, and the Team Alignment Map. We outline the design principles (among others) that should be followed for developing visual inquiry tools for other strategic management problems. We evaluate the design theory with an experiment in which groups of students were asked to develop visual inquiry tools. The results show that, when provided with the design theory, students designed higher-quality visual inquiry tools. Our study addresses the lack of guidance in the development of visual inquiry tools and the lack of methodological guidance in design science research on how to theorize and formalize knowledge across multiple projects. We provide a methodological process for analyzing multiple-project data by bridging methodological insights from design science research and qualitative methods from the social sciences.

Gerald C. (Jerry) Kane, Professor of IS and Faculty Director for the Edmund H. Shea Jr. Center for Entrepreneurship, Boston College’s Carroll School of Management

Wikipedia: The Canary in the Coal Mine for Online Collaboration. Many companies are increasingly seeking to use online communities for collaboration and innovation. Perhaps the most robust of these is Wikipedia, the open online encyclopedia that hosts millions of articles in hundreds of languages. Wikipedia preserves the entire history of collaboration that has taken place over the past 15 years, providing a remarkable environment for studying online collaboration and knowledge creation. While companies may not be interested in creating a Wikipedia-type platform, per se, the lessons learned in such a massive and robust online community can provide important insights for how to support innovation within companies (i.e. among employees) or across company boundaries (i.e. between employees and customers).

Merrill Warkentin, Professor of Information Systems, College of Business / Mississippi State University

Measurement Methods: Researching Individual Computer User Behaviors (with exemplars from security and privacy research). Recent attention has focused on ways to improve our methods for measuring individual computer user behaviors, including transitioning from “behavioral intention” (with surveys and experiments) to actual behavior measurements in various contexts, where possible. We will discuss this criticism and offer a number of innovative ways that computer user behaviors have been measured in the context of information security and privacy, including from my own recent research. This will include a discussion of lab experiments, field experiments, and neuro-physiological data collection methods and what they promise for improving our research metrics.

Roman Lukyanenko, Membre du GReSI et Professeur adjoint, Département des TI, HEC Montréal

Day Two on a Long and Treacherous Journey: Developing Definitions, Taxonomy and a Basic Ontology of Research Validities and Reliabilities. Validity and reliability are among the most widely used concepts in science. Broadly, both deal with the quality of scientific research and dependability of scientific findings. Many volumes have been written on the subject, with countless mentions and uses of the terms in scientific papers. Yet, confusion reigns supreme and we continue to lack established definitions and rigorous understanding of what reliability and validity means. With the help of modern tools for extraction and analysis of data, we are assembling the largest to date compendium of research validities and reliabilities (currently, over 400 distinct terms). Having obtained and curated this unique data source, we use it to develop definitions for both reliability and validity, definitions for each specific validity, and developing a basic ontology for research validities and reliabilities. The presentation will share some of the work in progress and preliminary findings.

Alina Dulipovici, Membre du GReSI et Professeure agrégée, Département des TI, HEC Montréal

Towards an Understanding of Information Security Behaviors: A Constructive Account of Organizational Values, Beliefs and Climate Characteristics. As digitalization shifts the focus from technology to information, it’s even more important for organizations to appropriately protect the confidentiality, integrity, and availability of organizational data. Security breaches caused by human error represent, according to several industry surveys, around a third of all security breaches and they are extremely damaging to organizations. Information security training and awareness programs seek to empower IT users to become security assets, but their influence on users’ behaviour is mitigated as IT users seem to have an irrational behaviour. Thus, we need more theorizing about the organizational context and the people interacting in the contexts we are theorizing about. To do so, this exploratory study draws on the theoretical frameworks of organizational climate (James and Jones 1974; Reichers and Schneider 1990) and organizational culture (Martin 1992; Schein 1984) that, together, influence users’ behaviours. By examining these concepts together (rather than separately as the extant research in information security does (e.g., Chan et al. 2005; Da Veiga and Martins 2015)), we provide a broader and more detailed view of the underlying forces influencing information security behaviors. These results represent a first step towards the development of a conceptually grounded instrument to measure organizational characteristics related to information security behaviors.

Camille Grange, Membre du GReSI et Professeure adjointe, Département des TI, HEC Montréal

A Methodology for Improving the Practical Relevance of Beliefs-Based Information Technology Usage Theories. There has been extensive research in Information Systems (IS) regarding the individual level beliefs that are germane to technology adoption and use. Theories such as the technology acceptance model, user satisfaction, service quality, the diffusion of innovations theory, and others identify a wide variety of influential beliefs, such as those that target properties of the system itself (e.g., its features), the use of it (e.g., its ease of use), or other contextual aspects (e.g., the pressure from peers to use it). These theories have been invaluable for understanding one of the most fundamental phenomena of interest to IS researchers and practitioners: individuals’ adoption and use of information technology (IT). However, as valuable as these theories are, there has been growing criticism about the fact that their constituting belief constructs (e.g., usefulness or quality) do not lend themselves to prescriptions for actionable interventions, in particular those geared towards IT design. We address this concern by developing a methodological approach to help researchers develop beliefs-based theories that are more relevant to IT practitioners. This approach relies on Fishbein and Ajzen’s (1975) foundational work on the nature and formation of individual beliefs to propose the integration and linking of constructs with different levels of broadness into beliefs-based IT usage theories. In order to illustrate the application of our approach and demonstrate its validity, we apply it to the context of e-business, decomposing the broad construct of supporting-service functionality, a strong predictor of satisfaction and continued usage, into two sets of narrower constructs that are more amenable to actionable recommendations for IT practitioners.

Hyung Koo Lee, Membre du GReSI et Professeur adjoint, Département des TI, HEC Montréal

Reporting Information Security Policy Violations – An exploratory study. Information security policy (ISP) violations have become a serious concern in organizations. Such violations need to be reported so that organizations can take immediate actions. In many cases, organizations rely on their employees to identify policy violations. However, prior research in this domain has focused primarily on individual and organizational factors that lead to violations of ISPs, and our current understanding of what influences individuals’ ISP violation reporting decisions is limited. In this study, we attempt to address this gap by conducting an explorative study on investigating the factors that drive ISP violation reporting.

Kai R. Larsen, Associate Professor of IS, Leeds School of Business / University of Colorado Boulder

Employing Natural Language Processing and Machine Learning to Improve Psychometrics: Can we Develop a Distance Measure for Constructs in IS, Management, and Psychology? Professor Larsen estimates the current size of the psychometrically-based literature to be at least three million articles. No one knows what the entirety of this literature contains, and we have no tools capable of integrating past findings about construct measurement item (survey questions) relationships across these articles. In this presentation, he will introduce a relatively simple and understandable process for the development of a consistent metric between construct measurement items. In this work, the individual survey item is represented by short L-vectors, and a unit of measurement, the L-measure, is specified as the average distance in space between an item and a set of other items where those items first start to approach a zero probability of measuring the same psychometric domain. The long-term goal of the research is to imbue all relevant items in a discipline with L-vector representations derived from high-resolution language spaces, and, consequently, reduced through supervised machine learning to only the dimensions of the vectors that drive empirical results between items. A probabilistic framework termed Revelation of Nomological Network (RONIN) is being developed for measurement of the structure of items and for reimagining construct validity evaluations. The presentation will focus on early results and visualization of item networks.

Marie-Claude Trudel, Membre du GReSI et Professeure agrégée, Département des TI, HEC Montréal

Extending the Theory of the Smart Machine to professional bureaucracies: The role of power. We extended one of the first native theories in information systems, Zuboff’s Theory of the Smart Machine, so that it can also be applied to professional bureaucracies operating in either for-profit or non-profit contexts. In our theoretical development, we took in consideration the specific context and power configuration of professional bureaucracies to 1) suggest the dilemma of power as a generalization of the dilemma of authority, 2) position the market as an external latent force of influence on innovating with information technology and 3) introduce external active forces of influence on innovating with information technology: government, professional associations and boards of directors. Our extended theory will help researchers and practitioners to better understand the elements that influence people to innovate fundamentally with information technology in professional bureaucracies. Keywords: Power; Professional bureaucracy; Theory of the Smart Machine; Theory extension; Information systems.

Placide Poba-Nzaou, Professeur en TI, Département d’organisation et ressources humaines, UQAM

Worries of open source projects’ contributors: Patterns, structures and engagement implications. Several organizations depend on open source software (OSS) code or talent but scholars have expressed concerns about OSS projects sustainability because of the challenges they face and their high dependency on volunteers’ contributions. Acknowledging that (1) challenges that contributors worry about may impact their level of engagement, (2) there have been few research on nonpathological worry with non-clinical samples in work-related context as opposed to the pathological and clinical focus (3) calls have been made for further empirical research on equifinality perspective at a micro- and meso-level (or group level) as opposed to organizational level, and from a behavioral perspective. This exploratory research uses factor and cluster analysis to derive empirical worrying profiles and examines profile differences in the relative level of relevant antecedents as well as equifinality in terms of contributors’ engagement (hours per week working on OSS projects). Based on an open dataset from a worldwide survey (N=1024), the study reveals the existence of four dimensions of OSS worry construct and five worrying profiles with marked configurational differences: no major worry (34.70%), major worry on software quality (19.30%), major worry on attraction and retention (18.60%), major worry on entrepreneurial responsibilities (12.40%), and major worry on communications (15%). Among the five profiles, two profiles exhibit equifinality with high (equifinal) engagement level.

Jim Denford, Professeur agrégé en SI et Doyen de la Faculté des sciences humaines et sociales, Collège militaire royal du Canada (CMR)

Une introduction à l’analyse quali-quantitative comparée (AQQC). Cette conférence de recherche portera sur l’analyse quali-quantitative comparée (AQQC) et sur les façons dont elle peut aider les chercheurs à élaborer et à tester de nouvelles théories plus riches. La relation étroite entre les technologies de l’information et les éléments organisationnels et environnementaux a créé un système complexe qui présente souvent des changements non linéaires et discontinus. Dans cette dynamique complexe, le rôle de l’informatique peut être mieux compris en tant qu’élément de l’ensemble du système plutôt qu’en tant que variable indépendante distincte. Récemment, l’AQQC s’est démarquée en raison de sa capacité à étudier les phénomènes complexes en utilisant une approche configurationnelle.