We study a data-driven bundle pricing problem with the following objective: given a predefined set of bundles, determine prices that maximize revenue under utility-maximizing consumer behavior. Classical mixed bundling models either assume a complete bundle menu with single-bundle choice or impose simplified pricing structures to ensure scalability. In contrast, we consider a setting in which the seller offers a restricted and potentially overlapping bundle catalog, allowing consumers to optimally select multiple non-overlapping bundles. This setting necessitates addressing two tightly coupled challenges: determining the price of each bundle and computing the revenue it generates under dependent consumer choice. Together, these challenges introduce additional combinatorial complexity in consumer decision-making and preserve strong price interdependencies across bundles. To address these challenges, we propose the Integrated Price Mining Method (IPMM), a scalable approach that approximates fully conditional pricing while explicitly accounting for dependent demand. We formalize consumer response as a utility maximization problem and show that it corresponds to a weighted set packing formulation. For price determination, IPMM combines balanced partitioning, sequential price search guided by competitive independence, dominance-based pruning, and cross-partition refinement to manage price dependence efficiently. Extensive experiments on synthetic instances and curated real-world bundle catalogs demonstrate that IPMM consistently achieves higher revenue coverage than classical and modern pricing baselines while remaining computationally tractable.
Organizations are beginning to deploy causal machine learning (causal ML) models for estimating average treatment effects and treatment effect differences for various subgroups, as well as identifying causal mechanisms driving treatment effects. For instance, in domains such as health care and marketing, causal ML is being used to identify who is most likely to benefit from or be receptive to treatments or interventions, including medical treatments and marketing campaigns. However, while causal ML models can produce highly accurate results when causal assumptions are met, the outputs can be confusing to work with. For instance, it is easy to conflate risk scores with conditional average treatment effects (CATEs), as risk scores from predictive models identify who is most likely to experience an outcome while CATEs from causal models identify who is most likely to benefit (or not benefit) from a treatment. A practitioner reviewing causal ML results may not fully understand such nuances and, thus, may not effectively convert relevant causal ML outputs into practice. Thus, a particular challenge relevant to causal ML, and extensible to any complex set of digital artifact outputs, is that effective practice transformation requires effective boundary spanning. To address this challenge, we propose a framework that conceptualizes large language models (LLMs) as dynamic boundary spanning objects. The framework leverages different forms of human-LLM collaborations throughout the causal ML boundary spanning process, toward the goal of successful practice transformation. We further provide an example of how the framework can be applied to a case of complex causal ML results. We ultimately seek to extend boundary spanning theory and provide a pathway for translating causal ML results across syntactic, semantic, and pragmatic boundaries.
This is an erratum for the article "AI Development and Innovation: A Comparison of Large Language Models from the U.S. and China" published in ACM Trans. Manag. Inform. Syst. 16, 4, Article 34 (November 2025), 18 pages.
Education program rankings have long been relied upon by students, parents, and institutions for their ability to assess academic quality and guide educational choices. While rankings significantly influence institutional strategies and student decision-making, traditional ranking systems often lack the nuanced insights needed for truly informed choices. To bridge this gap, this article introduces an information technology (IT) artifact that enables personalized educational program selection by matching students' career aspirations and skill development goals with relevant program offerings. We develop a three-component architecture comprising: a Wide & Deep skill extractor that identifies labor market skill requirements, CookieLM that generates comprehensive skill representations through language model and knowledge enhancement, and EPRQ that delivers personalized program recommendations aligned with individual career objectives. The empirical analyses demonstrate our framework's superior effectiveness over existing techniques and offer insightful, user-specific evaluations across different scenarios.
Recommendation algorithms are crucial to social media platforms, determining the order of information users see and aiding in discovering news. However, they can inadvertently cause harm through algorithmic privilege, excessive personalization, filter bubbles, and misinformation recommendation and propagation. Spreading misinformation through online social networks like Facebook, X (formerly Twitter), TikTok, and Instagram is commonplace in today’s milieu. Similar such cases were during the COVID-19 pandemic and the US Presidential election where misleading information spread widely. Such issues highlight the need for transparency and oversight in recommendation algorithms design. While some research has been conducted on mitigating misinformation within recommendation algorithms, the focus has primarily been on intervention strategies, which have been observed to cause a backfire effect. Likewise, the current literature only focuses on addressing misinformation at the time of recommendation, particularly without considering dissemination which is very concerning. Our study focuses on strategies to generate recommendations by modeling trustworthy social network and/or user neighbors aggregation while designing user-item latent factors to mitigate misinformation recommendation and dissemination. We evaluated various trustworthy user modeling designs on two Twitter datasets to assess the misinformation in top-K recommendation lists and simulated the propagation of recommended misinformation within the user network to also analyze misinformation spread. The research findings indicate that trustworthy social network and/or user neighbors modeling helps to mitigate misinformation spread at both at time of recommendation and dissemination.
