Recommendation systems have become ubiquitous tools in online platforms, providing personalized suggestions based on user–item interactions. To capture the dynamic higher-order connections between users and items, recommendation approaches based on dynamic graphs have garnered significant attention from researchers. However, existing recommendation methods based on dynamic graphs are often limited by data sparsity, which prevents them from achieving satisfactory performance. Fortunately, the rapid development of large language models (LLMs) with powerful text generation capabilities and extensive domain knowledge has offered new possibilities for addressing this challenge. However, how to effectively integrate LLMs with dynamic graphs remains unexplored. To bridge this gap, in this article, we propose a novel framework, that is, DynLLM, for applying LLMs to dynamic graph-based recommendation methods. Specifically, DynLLM harnesses the power of LLMs to generate multi-faceted user profiles based on the rich textual features of historical purchase records, which in turn supplement and enrich the underlying relationships between users and items. Along this line, to fuse the multi-faceted profiles with temporal graph embedding, we engage LLMs to derive corresponding profile embeddings and further employ a distilled attention mechanism to refine the LLM-generated profile embeddings for alleviating noisy signals, while also assessing and adjusting the relevance of each distilled facet embedding for seamless integration with temporal graph embedding from continuous time dynamic graphs (CTDGs). Extensive experiments on three real datasets have validated the superior improvements of DynLLM over a wide range of state-of-the-art baseline methods. The implementation code is available online at https://github.com/meteor-gif/DynLLM .

Last year has witnessed the re-flourishment of tag-aware recommender systems supported by the LLM-enriched tags. Unfortunately, though large efforts have been made, current solutions may fail to describe the diversity and uncertainty inherent in user preferences with only tag-driven profiles. Recently, with the development of geometry-based techniques, e.g., box embeddings, the diversity of user preferences now could be fully modeled as the range within a box in high dimension space. However, defect still exists as these approaches are incapable of capturing high-order neighbor signals, i.e., semantic-rich multi-hop relations within the user-tag-item tripartite graph, which severely limits the effectiveness of user modeling. To deal with this challenge, in this paper, we propose a novel framework, called BoxGNN, to perform message aggregation via combinations of logical operations, thereby incorporating high-order signals. Specifically, we first embed users, items, and tags as hyper-boxes rather than simple points in the representation space, and define two logical operations, i.e., union and intersection, to facilitate the subsequent process. Next, we perform the message aggregation mechanism via the combination of logical operations, to obtain the corresponding high-order box representations. Finally, we adopt a volume-based learning objective with Gumbel smoothing techniques to refine the representation of boxes. Extensive experiments on two publicly available datasets and one LLM-enhanced e-commerce dataset have validated the superiority of BoxGNN compared with various state-of-the-art baselines. The code is released online: https://github.com/critical88/BoxGNN.

Deep metric learning has been widely adopted to construct good representations for images and sentences with pair-based loss functions such as contrastive loss and triplet loss. However, these loss functions restrict the effectiveness of learned embedding and face two critical challenges: 1) high bias on account of a large set of uninformative and redundant pairs; 2) low information exploration due to the lack of processing the relation information among pairs. In this paper, we propose the jointly multi-similarity (JMS) loss to address the above challenges with a novel pair weighting strategy that assigns higher weights to the more informative pairs and discards the less informative ones. Specifically, we integrate the relationship information among pairs into a single framework and optimize the JMS loss by considering various information jointly. Furthermore, we extensively compare the JMS loss with other start-of-the-art approaches by conducting multiple experiments on image retrieval and semantic text similarity tasks. The experimental results show that the JMS loss consistently outperforms competitors.

Constrained action-based decision-making is one of the most challenging decision-making problems. It refers to a scenario where an agent takes action in an environment not only to maximize the expected cumulative reward but where it is subject to certain action-based constraints; for example, an upper limit on the total number of certain actions being carried out. In this work, we construct a general data-driven framework called Constrained Action-based Partially Observable Markov Decision Process (CAPOMDP) to induce effective pedagogical policies. Specifically, we induce two types of policies: CAPOMDPLG using learning gain as reward with the goal of improving students' learning performance, and CAPOMDPTime using time as reward for reducing students' time on task. The effectiveness of CAPOMDPLG is compared against a random yet reasonable policy and the effectiveness of CAPOMDPTime is compared against both a Deep Reinforcement Learning induced policy and a random policy. Empirical results show that there is an Aptitude-Treatment Interaction effect: students are split into High vs. Low based on their incoming competence; while no significant difference is found among the High incoming competence groups, for the Low groups, students following CAPOMDPTime indeed spent significantly less time than those using the two baseline policies and students following CAPOMDPLG significantly outperform their peers on both learning gain and learning efficiency.

Bayesian Knowledge Tracing (BKT) is one of the most widely adopted student-modeling methods. It uses performance (incorrect,correct) to infer student knowledge state (unlearned, learned). However, performance can be noisy and thus we explored another type of observations -- student response time. Furthermore, we proposed Intervention Bayesian Knowledge Tracing (Intervention-BKT) which can incorporate multiple types of instructional interventions into the conventional BKT model. Our results show that for next-step performance predictions, Intervention-BKT is more effective than BKT; whereas to predict students' post-test scores, including student response time would yield better result than using performance alone.

Reinforcement Learning (RL) is one of the best machine learning approaches for decision making in interactive environments. RL focuses on inducing effective decision making policies with the goal of maximizing the agent's cumulative reward. In this study, we investigated the impact of both immediate and delayed reward functions on RL-induced policies and empirically evaluated the effectiveness of induced policies within an Intelligent Tutoring System called Deep Thought. Moreover, we divided students into Fast and Slow learners based on their incoming competence as measured by their average response time on the initial tutorial level. Our results show that there was a significant interaction effect between the induced policies and the students' incoming competence. More specifically, Fast learners are less sensitive to learning environments in that they can learn equally well regardless of the pedagogical strategies employed by the tutor, but Slow learners benefit significantly more from effective pedagogical strategies than from ineffective ones. In fact, with effective pedagogical strategies the slow learners learned as much as their faster peers, but with ineffective pedagogical strategies the former learned significantly less than the latter.