Malek Ben Sliman: Homepage

Research Overview:

My research lies in the application of econometrics, machine learning and graph theory as well as analytical and empirical modeling tools to generate actionable insights for companies. For instance, I developed graph theoretical results to gain a better understanding of social networks and devised efficient stochastic seeding strategies in large or unobservable networks in order to leverage word-of-mouth and spread information and viral content. Additionally, I focused my dissertation on the drivers of art valuation and on the development of a model to predict future auction results by collecting and analyzing data regarding over 140,000 fine art auctions including the paintings’ images. In particular, I blend deep learning and unsupervised learning techniques, as well as econometrics approaches to assess the visual features, the level of creativity and influence of a painting and the evolution of artists’ reputations impacts on auction prices.

Working Papers:

Friendship Paradox Generalizations and Centrality Measures
Malek Ben Sliman and Rajeev Kohli
under review with Social Networks
[Show Abstract] [Working Paper (ResearchGate)] [Working Paper (SSRN)]

The friendship paradox says that on average your friends have more friends than you. Equivalently, in an undirected graph, the average degree of the nodes is no greater than the average degree of the terminal node of a random walk of length one. We generalize this result in two ways: (1) by considering directed graphs, which also allow one-sided relations, such as between followers and leaders; and (2) by characterizing the relations between the expected values of the in and out degrees of the terminal nodes of random alternating walks of length 2k and 2k+1, where k is nonnegative. The limiting value of these averages is proportional to the largest singular value of the associated adjacency matrix, and to its largest eigenvalue in the special case of an undirected graph. We interpret the results for one-sided relations (e.g., between leaders and followers) and two-sided relations (e.g., between friends). We further relate such extension to centrality measures. We show that beta centrality approaches eigenvector centrality when the inverse of the weighting parameter becomes arbitrarily close to the limiting value of the expected degree of the terminal node of a random walk of length k. When relations are asymmetric, this observation extends to the singular vectors of the associated directed graph.

Adaptive Customization
Malek Ben Sliman, Khaled Boughanmi and Rajeev Kohli
under review with Marketing Science.
[Show Abstract]

Most online retailers and content providers offer feature-based filtering tools to facilitate product search by their customers. We propose a method that learns about a customer's preferences each time he/she selects a screening feature, and then customizes the screening menu and displays the filtered alternatives in suitable order. It also allows a retailer to create adaptive displays showing the best alternatives for a highlighted feature. The proposed method has an offline component that estimates the parameters of a latent-class screening model, and an online component that uses Bayes' theorem to dynamically predict the sequential filtering of alternatives. We describe estimation procedures and illustrate adaptive customization using data from a choice experiment for electronic tablets.

R2M Index 1.0: Assessing the Relevance to Marketing of Academic Marketing Research
Kamel Jedidi, Bernd Schmitt, Malek Ben Sliman, and Yanyan Li.
Revision invited at Journal of Marketing
[Show Abstract]

Has marketing academia lost its relevance for marketing practice? To move beyond the long-lasting conceptual debate, this article assesses this question empirically by employing a text-mining methodology on academic marketing articles. We develop version 1.0 of the R2M (Relevance to Marketing) Index and validate it against external measures: the index correlates with relevance measures, such as practice awards and practitioner evaluations of articles, and articles in marketing journals have a higher score than articles in disciplinary journals in psychology and economics. Moreover, we find that articles published in the Journal of Marketing have the highest R2M score followed by Marketing Science, Journal of Marketing Research, and Journal of Consumer Research. We suggest that journal editors and authors use the R2M Index to assess the relevance of an article, and marketing practitioners to identify relevant articles. Because the index will always be a work in progress that requires continuous improvement, we discuss how to overcome limitations by including more journals, designing further safeguards against ``gaming the system,'' and moving from relevance to business impact.

Work in Progress:

Leveraging the Friendship Paradox for Seeding in Asymmetric Networks
Malek Ben Sliman and Rajeev Kohli
In preparation, to be submitted at Management Science.
[Show Abstract] [Working Paper (ResearchGate)] [Working Paper (SSRN)]

A critical assumption of the friendship paradox is the symmetry of the relationship. That is, if two people are friends, then each is the other's friend. In social networks like Facebook where this assumption is satisfied, the friendship paradox implies that firms can potentially achieve faster and more widespread diffusion of information by seeding it with the friends of a group of people than with people in the group itself. We generalize the result to allow one-sided (leader/follower) relations and examine the implications for seeding in social networks like Twitter where messages (such as tweets) can be transmitted in only one direction, from leaders to followers. We use the results to compare seeding with (1) leaders, (2) followers, and (3) ``friends,'' a sample of users selected by ignoring the asymmetry in relations. We obtain necessary and sufficient conditions under which each of three seeding methods obtains the highest number of followers per seed. We analyze a subset of Twitter data and find that seeding with leaders performs better than the other two methods.

The Art of Art Valuation
Malek Ben Sliman, Rajeev Kohli, and Kamel Jedidi
In preparation.
[Show Abstract]

Wealthy consumers increasingly invest in art which is considered as a more individualistic and exclusive form of luxury consumption. This trend is notably illustrated by the collaborations between luxury companies and famous artists. Yet, the drivers of the financial value of pieces of art are not fully understood. Indeed, even major auction houses are at awe when paintings are sold at prices multiple times higher than what they expected. The purpose of this paper is to deepen our understanding of the art world by investigating the drivers of art valuation. We collected details regarding over 140,000 fine art auction including the paintings’ images to address the following questions. First, we use deep learning tools to assess which visual features predict auction prices. Second, we create a network of paintings to establish how creative and influential each painting is. Finally, we shift our focus to artists and model the evolution of artists’ reputations over time.