Lukas De Kerpel
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Lukas De Kerpel has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Dries F. Benoit
Faculty of Economics and Business Administration, Ghent University, Ghent, Belgium
2 shared publications - Anneleen Van Kerckhove
Univ Ghent, Dept Mkt Innovat & Org, Tweekerkenstr 2, B-9000 Ghent, Belgium
1 shared publication - Arthur Thuy
Faculty of Economics and Business Administration, Ghent University, 9000 Ghent, Belgium; and Corelab CVAMO, FlandersMake@UGent, Ghent University, 9000 Ghent, Belgium
1 shared publication - Gudrun Roose
Ghent University
1 shared publication - Laura De Kerpel
Department of Marketing, Innovation, and Organisation, Faculty of Economics and Business Administration, Ghent University, Ghent, Belgium
1 shared publication
Latest Publications
- 1 May 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Transactions on Education
- Vol. 26 (3) , 187-199
A Bandit-Based Approach to Educational Recommender Systems: Contextual Thompson Sampling for Learner Skill Gain Optimization
In recent years, instructional practices in operations research, management science, and analytics have increasingly shifted toward digital environments, where large and diverse groups of learners make it difficult to provide practice that adapts to individual needs. This paper introduces a method that generates personalized sequences of exercises by selecting, at each step, the exercise most likely to advance a learner’s understanding of a targeted skill. The method uses information about the learner and their past performance to guide these choices, and learning progress is measured as the change in estimated skill level before and after each exercise. Using data from an online mathematics tutoring platform, we find that the approach recommends exercises associated with greater skill improvement and adapts effectively to differences across learners. From an instructional perspective, the framework enables personalized practice at scale, highlights exercises with consistently strong learning value, and helps instructors identify learners who may benefit from additional support.
- 1 May 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Transactions on Education
- Vol. 26 (3) , 187-199
A Bandit-Based Approach to Educational Recommender Systems: Contextual Thompson Sampling for Learner Skill Gain Optimization
In recent years, instructional practices in operations research, management science, and analytics have increasingly shifted toward digital environments, where large and diverse groups of learners make it difficult to provide practice that adapts to individual needs. This paper introduces a method that generates personalized sequences of exercises by selecting, at each step, the exercise most likely to advance a learner’s understanding of a targeted skill. The method uses information about the learner and their past performance to guide these choices, and learning progress is measured as the change in estimated skill level before and after each exercise. Using data from an online mathematics tutoring platform, we find that the approach recommends exercises associated with greater skill improvement and adapts effectively to differences across learners. From an instructional perspective, the framework enables personalized practice at scale, highlights exercises with consistently strong learning value, and helps instructors identify learners who may benefit from additional support.
- Open Access
- 5 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Recommender Systems
- Vol. 4 (3) , 1-23
A Reward-Informed Semi-Personalized Bandit Approach for Enhancing Accuracy and Serendipity in Online Slate Recommendations
Contextual bandits provide a principled framework for personalization in online recommendation settings. However, as these methods tailor recommendation slates to an individual user, they tend to induce overspecialization, yielding homogeneous recommendation lists that limit exposure to diverse content and contribute to more systemic issues such as filter bubbles and echo chambers. To mitigate these effects, recommender systems must complement predictive accuracy with serendipity, providing recommendations that are novel and unexpected while remaining contextually relevant. This study proposes a semi-personalized bandit that, for each item, learns a decision tree to segment users by contextual features and reward patterns, and runs a unique Thompson Sampling policy for each user segment to create recommendation slates. By pooling information across behaviorally similar users and conducting the exploration mechanism at the user segment level, the framework mitigates overspecialization issues and promotes serendipitous recommendations. Moreover, the approach is inherently interpretable, with decision trees revealing decision pathways that define user segments, offering insights into recommendation logic. Experiments across three different online domains show that the semi-personalized framework reduces average regret relative to personalized baselines while improving serendipity in sparse interaction settings. These findings underscore the potential of semi-personalized bandits to improve recommendation quality in complex environments.
- 5 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Recommender Systems
- Vol. 4 (3) , 1-23
Open AccessA Reward-Informed Semi-Personalized Bandit Approach for Enhancing Accuracy and Serendipity in Online Slate Recommendations
Contextual bandits provide a principled framework for personalization in online recommendation settings. However, as these methods tailor recommendation slates to an individual user, they tend to induce overspecialization, yielding homogeneous recommendation lists that limit exposure to diverse content and contribute to more systemic issues such as filter bubbles and echo chambers. To mitigate these effects, recommender systems must complement predictive accuracy with serendipity, providing recommendations that are novel and unexpected while remaining contextually relevant. This study proposes a semi-personalized bandit that, for each item, learns a decision tree to segment users by contextual features and reward patterns, and runs a unique Thompson Sampling policy for each user segment to create recommendation slates. By pooling information across behaviorally similar users and conducting the exploration mechanism at the user segment level, the framework mitigates overspecialization issues and promotes serendipitous recommendations. Moreover, the approach is inherently interpretable, with decision trees revealing decision pathways that define user segments, offering insights into recommendation logic. Experiments across three different online domains show that the semi-personalized framework reduces average regret relative to personalized baselines while improving serendipity in sparse interaction settings. These findings underscore the potential of semi-personalized bandits to improve recommendation quality in complex environments.
- 9 February 2026
- Published by SAGE Publications in Journal of Interactive Marketing
- Vol. 61 (3) , 291-310
Employee Influencers: Leveraging Employee Advocacy for Social Media Success
In today's digital landscape, organizations like to encourage employees to participate in employee advocacy by posting organizational content on their personal social media accounts. While engagement on official brand pages is often limited, content posted directly by employees typically generates stronger resonance with social media users. Employee advocacy thus involves a two-step process: Before content can reach and engage social media users, employees—acting as “gatekeepers”—must first post it within their own networks. Through two studies, this research explores how various visual-centric post characteristics influence both employees’ posting behavior and subsequent social media user engagement. Study 1 explores these effects within employee advocacy platforms, where organizations provide premade messages for employees to post, analyzing how different visual characteristics impact employee postings and social media user clicks. Study 2 focuses on LinkedIn, where content posted by employees may originate from the organization or be independently created. Using negative binomial regressions, both studies analyze which visual characteristics—such as format or image type—drive employee postings (Study 1 and 2), social media user clicks (Study 1), and social media user reactions (Study 2). The findings highlight key similarities and differences in preferences, offering actionable insights for optimizing employee advocacy strategies.
- 9 February 2026
- Published by SAGE Publications in Journal of Interactive Marketing
- Vol. 61 (3) , 291-310
Employee Influencers: Leveraging Employee Advocacy for Social Media Success
In today's digital landscape, organizations like to encourage employees to participate in employee advocacy by posting organizational content on their personal social media accounts. While engagement on official brand pages is often limited, content posted directly by employees typically generates stronger resonance with social media users. Employee advocacy thus involves a two-step process: Before content can reach and engage social media users, employees—acting as “gatekeepers”—must first post it within their own networks. Through two studies, this research explores how various visual-centric post characteristics influence both employees’ posting behavior and subsequent social media user engagement. Study 1 explores these effects within employee advocacy platforms, where organizations provide premade messages for employees to post, analyzing how different visual characteristics impact employee postings and social media user clicks. Study 2 focuses on LinkedIn, where content posted by employees may originate from the organization or be independently created. Using negative binomial regressions, both studies analyze which visual characteristics—such as format or image type—drive employee postings (Study 1 and 2), social media user clicks (Study 1), and social media user reactions (Study 2). The findings highlight key similarities and differences in preferences, offering actionable insights for optimizing employee advocacy strategies.