HEC Montréal

About: HEC Montréal is a education organization based out in Montreal, Quebec, Canada. It is known for research contribution in the topics: Vehicle routing problem & Corporate governance. The organization has 1221 authors who have published 5708 publications receiving 196862 citations. The organization is also known as: Ecole des Hautes Etudes Commerciales de Montreal & HEC Montreal.

Towards a precise definition of the OMG/MDA framework

Abstract: We are currently witnessing an important paradigm shift in information system construction, namely the move from object and component technology to model technology. The object technology revolution has allowed the replacement of the over twenty-year-old step-wise procedural decomposition paradigm with the more fashionable object composition paradigm. Surprisingly, this evolution seems to have triggered another even more radical change, the current trend toward model transformation. A concrete example is the Object Management Group's rapid move from its previous Object Management Architecture vision to the latest Model-Driven Architecture. This paper proposes an interpretation of this evolution through abstract investigation. In order to stay as language-independent as possible, we have employed the neutral formalism of Sowa's conceptual graphs to describe the various situations characterizing this organization. This will allow us to identify potential problems in the proposed modeling framework and suggest some possible solutions.

A comprehensive model of customer direct and indirect revenge: understanding the effects of perceived greed and customer power

Abstract: This article develops and tests a comprehensive model of customer revenge that contributes to the literature in three manners. First, we identify the key role played by the customer’s perception of a firm’s greed—that is, an inferred negative motive about a firm’s opportunistic intent—that dangerously energizes customer revenge. Perceived greed is found as the most influential cognition that leads to a customer desire for revenge, even after accounting for well studied cognitions (i.e., fairness and blame) in the service literature. Second, we make a critical distinction between direct and indirect acts of revenge because these sets of behaviors have different repercussions—in “face-to-face” vs. “behind a firm’s back”—that call for different interventions. Third, our extended model specifies the role of customer perceived power in predicting these types of behaviors. We find that power is instrumental—both as main and moderation effects—only in the case of direct acts of revenge (i.e., aggression and vindictive complaining). Power does not influence indirect revenge, however. Our model is tested with two field studies: (1) a study examining online public complaining, and (2) a multi-stage study performed after a service failure.

InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization

Abstract: This paper studies learning the representations of whole graphs in both unsupervised and semi-supervised scenarios. Graph-level representations are critical in a variety of real-world applications such as predicting the properties of molecules and community analysis in social networks. Traditional graph kernel based methods are simple, yet effective for obtaining fixed-length representations for graphs but they suffer from poor generalization due to hand-crafted designs. There are also some recent methods based on language models (e.g. graph2vec) but they tend to only consider certain substructures (e.g. subtrees) as graph representatives. Inspired by recent progress of unsupervised representation learning, in this paper we proposed a novel method called InfoGraph for learning graph-level representations. We maximize the mutual information between the graph-level representation and the representations of substructures of different scales (e.g., nodes, edges, triangles). By doing so, the graph-level representations encode aspects of the data that are shared across different scales of substructures. Furthermore, we further propose InfoGraph*, an extension of InfoGraph for semi-supervised scenarios. InfoGraph* maximizes the mutual information between unsupervised graph representations learned by InfoGraph and the representations learned by existing supervised methods. As a result, the supervised encoder learns from unlabeled data while preserving the latent semantic space favored by the current supervised task. Experimental results on the tasks of graph classification and molecular property prediction show that InfoGraph is superior to state-of-the-art baselines and InfoGraph* can achieve performance competitive with state-of-the-art semi-supervised models.