Recommender systems rely on Collaborative Filtering (CF) to predict user preferences by leveraging patterns in historical user-item interactions. While traditional CF methods primarily focus on learning compact vector embeddings for users and items, graph neural network (GNN)-based approaches have emerged as a powerful alternative, utilizing the structure of user-item interaction graphs to enhance recommendation accuracy. However, existing GNN-based models, such as LightGCN and UltraGCN, often struggle with two major limitations: an inability to fully account for heterophilic interactions, where users engage with diverse item categories, and the over-smoothing problem in multi-layer GNNs, which hinders their ability to model complex, high-order relationships. To address these gaps, we introduce WaveHDNN, an innovative wavelet-enhanced hypergraph diffusion framework. WaveHDNN integrates a Heterophily-aware Collaborative Encoder, designed to capture user-item interactions across diverse categories, with a Multi-scale Group-wise Structure Encoder, which leverages wavelet transforms to effectively model localized graph structures. Additionally, cross-view contrastive learning is employed to maintain robust and consistent representations. Experiments on benchmark datasets validate the efficacy of WaveHDNN, demonstrating its superior ability to capture both heterophilic and localized structural information, leading to improved recommendation performance.

Meta is giving its AI assistant a better "memory" in an effort to make the chatbot more useful. The company's latest AI update allows the assistant to "remember certain details that you share with it in 1:1 chat" and uses your past activity on Facebook and Instagram to make more personalized recommendations. With the change, which will initially be available to the US and Canada, Meta AI will be able to track your preferences based on information you share in chat with it. In a blog post, the company uses the example of food allergies and other dietary restrictions so meta AI will "remember" to recommend recipes that fit your requirements. But the assistant will also be able to track other details about you, including information about your personal life and relationships.

The authors give a statement that "the recommendation model is optimized over the worst-case exposure mechanism" but fail to give clear motivation behind the model. Why optimizing with the worst-case exposure is better than optimizing with the expected exposure that is widely adopted by existing methods? It seems that the essential advantage of the proposed method is robust. Uncertainty is not a good motivation as it has been considered by existing methods and can not answer the above question. The proposed method should be compared with existing unbiased recommendation methods (e.g. The difference in terms of solutions and generalization bounds between the paper with [a5][a6] should be discussed.

Reviews were quite borderline, but ultimately slightly on the positive side. Given the borderline scores, a discussion was initiated. The reviewers raised some issues about novelty/comparisons (R1,R3), motivation (R2), missing work (R2,R3), and experimental analysis (R4). Mostly though the reviewers did not consider these to be critical issues. The discussion eventually resulted in some positive movement of the scores/comments and reached a consensus around recommending acceptance.

Feature modeling, which involves feature representation learning and leveraging, plays an essential role in industrial recommendation systems. However, the data distribution in real-world applications usually follows a highly skewed long-tail pattern due to the popularity bias, which easily leads to over-reliance on ID-based features, such as user/item IDs and ID sequences of interactions. Such over-reliance makes it hard for models to learn features comprehensively, especially for those non-ID meta features, e.g., user/item characteristics. Further, it limits the feature leveraging ability in models, getting less generalized and more susceptible to data noise. Previous studies on feature modeling focus on feature extraction and interaction, hardly noticing the problems brought about by the long-tail data distribution. To achieve better feature representation learning and leveraging on real-world data, we propose a model-agnostic framework AdaF^2M^2, short for Adaptive Feature Modeling with Feature Mask. The feature-mask mechanism helps comprehensive feature learning via multi-forward training with augmented samples, while the adapter applies adaptive weights on features responsive to different user/item states. By arming base models with AdaF^2M^2, we conduct online A/B tests on multiple recommendation scenarios, obtaining +1.37% and +1.89% cumulative improvements on user active days and app duration respectively. Besides, the extended offline experiments on different models show improvements as well. AdaF$^2$M$^2$ has been widely deployed on both retrieval and ranking tasks in multiple applications of Douyin Group, indicating its superior effectiveness and universality.

User interests manifest a dynamic pattern within the course of a day, e.g., a user usually favors soft music at 8 a.m. but may turn to ambient music at 10 p.m. To model dynamic interests in a day, hour embedding is widely used in traditional daily-trained industrial recommendation systems. However, its discreteness can cause periodical online patterns and instability in recent streaming recommendation systems. Recently, Interest Clock has achieved remarkable performance in streaming recommendation systems. Nevertheless, it models users' dynamic interests in a coarse-grained manner, merely encoding users' discrete interests of 24 hours from short-term behaviors. In this paper, we propose a fine-grained method for perceiving time information for streaming recommendation systems, named Long-term Interest Clock (LIC). The key idea of LIC is adaptively calculating current user interests by taking into consideration the relevance of long-term behaviors around current time (e.g., 8 a.m.) given a candidate item. LIC consists of two modules: (1) Clock-GSU retrieves a sub-sequence by searching through long-term behaviors, using query information from a candidate item and current time, (2) Clock-ESU employs a time-gap-aware attention mechanism to aggregate sub-sequence with the candidate item. With Clock-GSU and Clock-ESU, LIC is capable of capturing users' dynamic fine-grained interests from long-term behaviors. We conduct online A/B tests, obtaining +0.122% improvements on user active days. Besides, the extended offline experiments show improvements as well. Long-term Interest Clock has been integrated into Douyin Music App's recommendation system.

The distributionally robust optimization (DRO)-based graph neural network methods improve recommendation systems' out-of-distribution (OOD) generalization by optimizing the model's worst-case performance. However, these studies fail to consider the impact of noisy samples in the training data, which results in diminished generalization capabilities and lower accuracy. Through experimental and theoretical analysis, this paper reveals that current DRO-based graph recommendation methods assign greater weight to noise distribution, leading to model parameter learning being dominated by it. When the model overly focuses on fitting noise samples in the training data, it may learn irrelevant or meaningless features that cannot be generalized to OOD data. To address this challenge, we design a Distributionally Robust Graph model for OOD recommendation (DRGO). Specifically, our method first employs a simple and effective diffusion paradigm to alleviate the noisy effect in the latent space. Additionally, an entropy regularization term is introduced in the DRO objective function to avoid extreme sample weights in the worst-case distribution. Finally, we provide a theoretical proof of the generalization error bound of DRGO as well as a theoretical analysis of how our approach mitigates noisy sample effects, which helps to better understand the proposed framework from a theoretical perspective. We conduct extensive experiments on four datasets to evaluate the effectiveness of our framework against three typical distribution shifts, and the results demonstrate its superiority in both independently and identically distributed distributions (IID) and OOD.

Weaknesses: 1) The experiment is somewhat inadequate. In the paper, the author only compares the proposed SIRI approach to the baseline from original Touchdown dataset paper [2]. In fact, spatial description resolution is a similar task as referring expression or instruction grounding. It is necessary for the author to further compare to approaches (such as Mattnet [18] or other new methods in 2019) in those tasks. For example, although Mattnet is not designed for spatial description resolution, but there is also semantic and position modules to handle spatial relation and object relationship reasoning, which can be served as a substitute of Part I & II of SIRI.

Paper was reviewed by four expert reviewers, with initial scores of: 6, 6, 6, 5. Reviewers acknowledge a commendable improvements of the proposed approach on a difficult and novel task. A number of issues where raised about the paper, including (1) poor exposition and language [all reviewers], (2) lack of comparison to FiLM [R1], (3) specificity of task and dataset [R2,R4], among others. Authors provided a rebuttal that was discussed by reviewers and ultimately convincing. Two of the reviewers upgraded their scores, resulting in unanimously positive, albeit marginally so, scores of: 7, 6, 6, 6. AC, despite having reservations about quality of the writing, mentioned by all reviewers, agrees that the approach is valuable and presents a significant improvement over state-of-the-art on a relatively unexplored problem. As such AC agrees with reviewers that the paper should be accepted.

The paper presents a novel method for recommendation with collaborative filtering based on Markov Random Fields (MRF). Starting from a general approach that regresses the full graph of items, the paper shows that a valid approximation can be obtained by proceeding with subgraphs that represent Markov blankets of an initial set of items. This approach yields significant computing gains, while yielding better recommendation performance compared to the state-of-the-art represented here by variational auto-encoders. As a general comment, I am wondering whether taking into account the popularity bias makes sense in the approach and if the authors thought about it. The claims are well supported by theoretical analysis.