Cette étude applique une nouvelle méthode de modélisation de sujet pour cartographier le contenu thématique du livre blanc des offres initiales de pièces de monnaie (ICO) afin d’analyser sa valeur d’information pour les investisseurs. À l’aide d’un algorithme de modélisation de sujets basé sur des phrases, nous déterminons et quantifions empiriquement 30 sujets dans une vaste collection de 5 210 livres blancs ICO entre 2015 et 2021. Nous constatons que l’algorithme produit un ensemble sémantiquement significatif de sujets, ce qui améliore considérablement les performances du modèle dans identifier les projets réussis. Les sujets les plus pertinents concernent les caractéristiques techniques de l’ICO.
- 00:03:47
Robotic warehouses have transformed logistics, prioritizing speed and efficiency. However, traditional static priority systems often leave low-priority customers facing excessive delays, raising concerns about fairness. This research, based on Invia, a robotic warehouse company, proposes a dynamic priority allocation model to balance efficiency and fairness. By adjusting order priorities over time, this approach ensures that both high-priority and long-waiting low-priority orders receive timely fulfillment. Through stochastic modeling and simulations, we demonstrate that dynamic prioritization reduces delays compared to static and first-come, first-served (FCFS) models. Case studies in e-commerce and healthcare logistics illustrate the broader impact of fairness in automation. As industries increasingly rely on AI-driven decision-making, the balance between efficiency and equity becomes critical. This research challenges the assumption that robotic warehouses should optimize for speed alone and advocates for a future where fairness plays a central role in automated commerce.
- Recherche
- Logistique et Supply Chain, Transformation Digitale
