Graph neural network-based recommenders employ the aggregation paradigms to learn node representation from higher-order neighboring nodes within the graph. However, these simple aggregation paradigms may perform poorly when mitigating noise impacts and capturing complex user preferences. To address it, some studies have attempted to enhance representation through contrastive augmentation across different views. Despite some effectiveness, the simple-view contrasts are still suboptimal with some unresolved challenges: (1) the influence of multivariate noise in interaction data, (2) knowledge biases introduced by irrelevant connections, and (3) user’s multiple interests. In this work, we propose a novel method named Diffusion-Based Relation Augmentation for Knowledge-aware Recommendation (DRAR) to overcome the above challenges. First, we alleviate the impact of interaction noise by injecting uncertainty and generating preference distributions with a diffusion-based module. Next, we design a relation augmentation module to effectively capture user neighborhood-level and context-level enhanced representations to alleviate the knowledge bias of irrelevant connections. Furthermore, we design a collaborative alignment module that enhances the model’s robustness by aligning user representation views at different stages. Extensive experiments on three benchmark datasets consistently demonstrate the superiority of our model over the state-of-the-art approaches. Our model demonstrates average improvements of 6.78% in Recall and 7.38% in NDCG across all datasets.

Click-Through Rate (CTR) prediction, which estimates the probability of a user clicking on a given item, is a critical task for online information services. Existing approaches often make strong assumptions that training and test data come from the same distribution. However, the data distribution varies since user interests are constantly evolving, resulting in the Out-of-Distribution (OOD) issue. In addition, users tend to have multiple interests, some of which evolve faster than others. Toward this end, we propose Disentangled Click-Through Rate Prediction (DiseCTR), which introduces a causal perspective of recommendation and disentangles multiple aspects of user interests to alleviate the OOD issue in recommendation. We conduct a causal factorization of CTR prediction involving user interest, exposure model, and click model, based on which we develop a deep learning implementation for these three causal mechanisms. Specifically, we first design an interest encoder with sparse attention which maps raw features to user interests and then introduce a weakly supervised interest disentangler to learn independent interest embeddings, which are further integrated by an attentive interest aggregator for prediction. Experimental results on three real-world datasets show that DiseCTR achieves the best accuracy and robustness in OOD recommendation against state-of-the-art approaches, significantly improving AUC and GAUC by over 0.02 and reducing logloss by over 13.7%. Further analyses demonstrate that DiseCTR successfully disentangles user interests, which is the key to OOD generalization for CTR prediction. We have released the code and data at https://github.com/DavyMorgan/DiseCTR/ .

The pervasive use of short-video applications has raised concerns about their potential negative effects on users, particularly addiction. Existing research often relies on psychological questionnaires, which lack real-world behavioral data, limiting scalability and analytical depth. To address this, we assess the addiction status of short-video platform users using a standardized psychometric questionnaire, combined with platform behavioral data and interview responses to uncover features associated with addiction. Using feature-based modeling, we scale to a dataset of 10,111 addiction-labeled users and identify key indicators of addiction, including prolonged daily watch time, especially at night, and excessive video consumption, while also revealing that higher watch frequency is not fully correlated with addiction. Additionally, we find that addicted users tend to consume a narrower range of content, suggesting a filter bubble effect. Our large-scale analysis provides valuable insights for platform designers, policymakers, and mental health professionals seeking to promote healthier engagement and mitigate the risks of short-video addiction.

Graph-based Cognitive Diagnosis (CD) has attracted much research interest due to its strong ability on inferring students' proficiency levels on knowledge concepts. While graph-based CD models have demonstrated remarkable performance, we contend that they still cannot achieve optimal performance due to the neglect of edge heterogeneity and uncertainty. Edges involve both correct and incorrect response logs, indicating heterogeneity. Meanwhile, a response log can have uncertain semantic meanings, e.g., a correct log can indicate true mastery or fortunate guessing, and a wrong log can indicate a lack of understanding or a careless mistake. In this paper, we propose an Informative Semantic-aware Graph-based Cognitive Diagnosis model (ISG-CD), which focuses on how to utilize the heterogeneous graph in CD and minimize effects of uncertain edges. Specifically, to explore heterogeneity, we propose a semantic-aware graph neural networks based CD model. To minimize effects of edge uncertainty, we propose an Informative Edge Differentiation layer from an information bottleneck perspective, which suggests keeping a minimal yet sufficient reliable graph for CD in an unsupervised way. We formulate this process as maximizing mutual information between the reliable graph and response logs, while minimizing mutual information between the reliable graph and the original graph. After that, we prove that mutual information maximization can be theoretically converted to the classic binary cross entropy loss function, while minimizing mutual information can be realized by the Hilbert-Schmidt Independence Criterion.Finally, we adopt an alternating training strategy for optimizing learnable parameters of both the semantic-aware graph neural networks based CD model and the edge differentiation layer. Extensive experiments on three real-world datasets have demonstrated the effectiveness of ISG-CD.

Dynamic link prediction aims to predict the future links on dynamic graphs, which can be applied to wide scenarios such as recommender systems and social networks on the World Wide Web. Existing methods mainly (1) focus on the in-graph learning, which cannot generalize to graphs unobserved during training; or (2) achieve the cross-graph predictions in a many-many mechanism by training on multiple graphs across various domains, which results in a large computational cost. In this paper, we propose a cross-graph dynamic link predictor named CrossDyG, which achieves the cross-graph transferability in a one-many mechanism which trains on one single source graph and test on different target graphs. Specifically, we provide causal and empirical analysis on the structural bias caused by the graph-specific structural characteristics in cross-graph predictions. Then, we conduct deconfounded training to learn the universal network evolution pattern from one single source graph during training. Finally, we apply the causal intervention to leverage the graph-specific structural characteristics of each target graph during inference. Extensive experiments conducted on three benchmark data of dynamic graphs demonstrate that CrossDyG outperforms the state-of-the-art baselines by up to 11.01% and 17.02% in terms of AP and AUC, respectively. In addition, the improvements are especially significant when training on small source graphs.

Recent advances in the large language models (LLM) have empowered traditional bots to gain human-level intelligence and exhibit human-like social behaviors, giving rise to a new form of LLM-driven social agents. However, the inherent limitations in LLMs could potentially result in politically biased behaviors of these agents, posing unexpected risks to human society. While great efforts have been made to examine political bias and related concerns in traditional bots and LLMs, little is known about the existence, unique characteristics, underlying origins, and potential mitigation strategies of this bias in LLM-driven social agents. To address this gap, we systematically assess political bias in LLM-driven social agents, by examining how it emerges as these agents self-reflect, communicate, and understand others during social interactions. Through designing and implementing social experiments, we discover that this bias consistently manifests in the social behaviors of agents driven by diverse LLMs, across nine key political topics. Inspired by the social learning theory, we propose to mitigate political bias by guiding these agents to emulate how humans learn to behave. By incorporating self-regulated and role-model learning processes, we reduce their political bias by 4.89% to 51.26% across diverse LLMs and topics, demonstrating the effectiveness and generalizability of the proposed strategy. This study not only advances the understanding of political bias in emerging LLM-driven agents, but also offers insights into harnessing social bots for social good.

In the past few years, infectious diseases like COVID-19 have caused serious distress to the global society and the economy. To prevent its spread, the early detection and assessment of infectious diseases based on molecular tests or antigen testing of bodily have led to countless labor and material costs. Fortunately, with the rapid development of mobile localization and web techniques, the collected massive mobile trajectory data provide a promising solution for detecting positive cases. However, existing mobility data-driven infection case detection methods are limited in terms of modeling the complicated epidemic spreading processes and preserving user privacy of the mobility data. In this article, we propose a novel graph convolutional networks (GCN) model for detecting high-risk infection cases, where we incorporate a spatio-temporal hypergraph to model the complex interaction of individuals. Then, we elaborately design a privacy-preserving framework tightly coupled with the structure of the spatio-temporal hypergraph, which includes a mobility data obfuscation module to protect privacy and an accompanying confidence-aware mechanism to mitigate the consequent performance decline. Moreover, we introduce a causal propagation mechanism to further guarantee the temporal dependency and causal effect of the feature propagation in our spatio-temporal hypergraph, which introduces both the causal transform of node features and the causal gathering of edge features. Finally, extensive experiments on a large mobility dataset collected from location-based services (LBS) show that the proposed model improves the performance of infection case detection by at least 12.47% when compared with several widely adopted baselines. Besides, our code and datasets are available at the link ( https://github.com/wjfu99/EPI-HGNN ).

Large language models (LLMs) have recently garnered significant attention in various domains, including recommendation systems. Recent research leverages the capabilities of LLMs to improve the performance and user modeling aspects of recommender systems. These studies primarily focus on utilizing LLMs to interpret textual data in recommendation tasks. However, it's worth noting that in ID-based recommendations, textual data is absent, and only ID data is available. The untapped potential of LLMs for ID data within the ID-based recommendation paradigm remains relatively unexplored. To this end, we introduce a pioneering approach called “LLM for ID-based recommendation” (LLM4IDRec). This innovative approach integrates the capabilities of LLMs while exclusively relying on ID data, thus diverging from the previous reliance on textual data. The basic idea of LLM4IDRec is that by employing LLM to augment ID data, if augmented ID data can improve recommendation performance, it demonstrates the ability of LLM to interpret ID data effectively, exploring an innovative way for the integration of LLM in ID-based recommendation. Specifically, we first define a prompt template to enhance LLM's ability to comprehend ID data and the ID-based recommendation task. Next, during the process of generating training data using this prompt template, we develop two efficient methods to capture both the local and global structure of ID data. We feed this generated training data into the LLM and employ LoRA for fine-tuning LLM. Following the fine-tuning phase, we utilize the fine-tuned LLM to generate ID data that aligns with users’ preferences. We design two filtering strategies to eliminate invalid generated data. Thirdly, we can merge the original ID data with the generated ID data, creating augmented data. Finally, we input this augmented data into the existing ID-based recommendation models without any modifications to the recommendation model itself. We evaluate the effectiveness of our LLM4IDRec approach using three widely used datasets. Our results demonstrate a notable improvement in recommendation performance, with our approach consistently outperforming existing methods in ID-based recommendation by solely augmenting input data.

Dynamic graph learning on temporal networks aims to understand the continuous evolution pattern of networks, with an important application on forecasting the future temporal network. Existing methods mainly focus on modeling the structural and temporal features, with recent research interest shifting toward considering the structural correlations between nodes through their neighbor co-occurrences. Though satisfactory performance has been achieved, there still remain several limitations: (1) the deviation of investigated scenarios from real-world applications, since most previous researches concentrate on special cases of multigraphs with abundant repeat edges; (2) the insufficient computational efficiency of modeling the structural features, since the existing neighbor co-occurrence scheme fails to consider explicit structural correlations between nodes and suffers from a time-consuming pairwise encoding strategy; (3) the unsatisfying prediction accuracy due to inadequate modeling of temporal features, since each neighbor’s historical temporal features and the temporal domain shifting with network evolving are both neglected. To solve these issues, we first focus on the general scenarios of temporal networks without abundant repeat edges for approaching the actual applications and propose an efficient and effective dynamic graph learning method named LightDyG. Specifically, (1) on the one hand, to increase the computational efficiency, LightDyG decouples the structural correlations between nodes and their individual substructures for fast convergence based on the analysis of existing co-occurrence mechanism, and further designs an incremental strategy for efficient structural encoding; (2) on the other hand, to improve the prediction accuracy, the temporal characteristics are considered by including both the interaction and appearance timestamps of neighbors, and a time-invariant temporal encoding strategy is designed to eliminate the temporal bias introduced by the network evolution. Extensive experiments conducted on four public temporal networks demonstrate that LightDyG outperforms the best baselines by 4.54–11.39% and 6.06–16.24% in terms of AP and AUC on the temporal link prediction tasks, respectively. In addition, LightDyG reduces the time cost for training and test up to 45.91% and 63.94%, respectively, and also achieves a fast convergence speed during training. The implementation of our approach is available in https://github.com/nudtzpan/LightDyG .

The mainstream approach of GNN-based recommendation aggregates high-order ID information associated with the node in the user-item graph. The aggregation pattern using ID as signal has two disadvantages: lack of textual semantics and the impact of interaction noise. These disadvantages pose a threat to effectively learn user preferences, especially in capturing intricate user-item semantic relationships. Although large language models (LLMs) allow the integration of rich textual information into recommenders and have had groundbreaking applications in recommender systems, current works need to bridge the gap between different representation spaces. This is because LLM-based methods align the representations of GNN-based models only by using text embedding of LLM, leading to unsatisfactory results. To address this challenge, we propose a denoising alignment framework with LLMs for GNN-based recommenders (DALR) , which aims to align structural representation with textual representation and mitigate the effects of noise. Specifically, we propose a modeling framework that integrates the representation of graph structure with textual information from LLMs to capture intricate user-item interactions. We also suggest an alignment paradigm to enhance representation performance by aligning semantic signals from LLMs and structural features from GNN models. Additionally, we introduce a contrastive learning scheme to relieve the impact of noise and improve model performance. Extensive experiments on public datasets demonstrate that our model consistently outperforms the state-of-the-art methods. DALR achieves improvements ranging from 2.82% to 12.20% in Recall@5 and from 1.04% to 3.48% in NDCG@5 compared to the strongest baseline model, using the Steam dataset as an example.

Session-based recommendation (SBR) methods often rely on user behavior data, which can struggle with the sparsity of session data, limiting performance. Researchers have identified that beyond behavioral signals, rich semantic information in item descriptions is crucial for capturing hidden user intent. While Large Language Models (LLMs) offer new ways to leverage this semantic data, the challenges of session anonymity, short-sequence nature, and high LLM training costs have hindered the development of a lightweight, efficient LLM framework for SBR. To address the above challenges, we propose an LLM-enhanced SBR framework that integrates semantic and behavioral signals from multiple views. This two-stage framework leverages the strengths of both LLMs and traditional SBR models while minimizing training costs. In the first stage, we use multi-view prompts to infer latent user intentions at the session semantic level, supported by an intent localization module to alleviate LLM hallucinations. In the second stage, we align and unify these semantic inferences with behavioral representations, effectively merging insights from both large and small models. Extensive experiments on two real datasets demonstrate that the LLM4SBR framework can effectively improve model performance. We release our codes along with the baselines at https://github.com/tsinghua-fib-lab/LLM4SBR .

Stance detection automatically detects the stance in a text towards a target, vital for content analysis in web and social media research. Despite their promising capabilities, LLMs encounter challenges when directly applied to stance detection. First, stance detection demands multi-aspect knowledge, from deciphering event-related terminologies to understanding the expression styles in social media platforms. Second, stance detection requires advanced reasoning to infer authors' implicit viewpoints, as stances are often subtly embedded rather than overtly stated in the text. To address these challenges, we design a three-stage framework COLA (short for Collaborative rOle-infused LLM-based Agents) in which LLMs are designated distinct roles, creating a collaborative system where each role contributes uniquely. Initially, in the multidimensional text analysis stage, we configure the LLMs to act as a linguistic expert, a domain specialist, and a social media veteran to get a multifaceted analysis of texts, thus overcoming the first challenge. Next, in the reasoning-enhanced debating stage, for each potential stance, we designate a specific LLM-based agent to advocate for it, guiding the LLM to detect logical connections between text features and stance, tackling the second challenge. Finally, in the stance conclusion stage, a final decision maker agent consolidates prior insights to determine the stance. Our approach avoids extra annotated data and model training and is highly usable. We achieve state-of-the-art performance across multiple datasets. Ablation studies validate the effectiveness of each role design in handling stance detection. Further experiments have demonstrated the explainability and the versatility of our approach. Our approach excels in usability, accuracy, effectiveness, explainability and versatility, highlighting its value.

Recommender systems filter out information that meets user interests. However, users may be tired of the recommendations that are too similar to the content they have been exposed to in a short historical period, which is the so-called user fatigue. Despite the significance for a better user experience, user fatigue is seldom explored by existing recommenders. In fact, there are three main challenges to be addressed for modeling user fatigue, including what features support it, how it influences user interests, and how its explicit signals are obtained. In this paper, we propose to model user Fatigue in interest learning for sequential Recommendations (FRec). To address the first challenge, based on a multi-interest framework, we connect the target item with historical items and construct an interest-aware similarity matrix as features to support fatigue modeling. Regarding the second challenge, built upon feature cross, we propose a fatigue-enhanced multi-interest fusion to capture long-term interest. In addition, we develop a fatigue-gated recurrent unit for short-term interest learning, with temporal fatigue representations as important inputs for constructing update and reset gates. For the last challenge, we propose a novel sequence augmentation to obtain explicit fatigue signals for contrastive learning. We conduct extensive experiments on real-world datasets, including two public datasets and one large-scale industrial dataset. Experimental results show that FRec can improve AUC and GAUC up to 0.026 and 0.019 compared with state-of-the-art models, respectively. Moreover, large-scale online experiments demonstrate the effectiveness of FRec for fatigue reduction. Our codes are released at https://github.com/tsinghua-fib-lab/SIGIR24-FRec.

In modern recommender systems, sequential recommendation leverages chronological user behaviors to make effective next-item suggestions, which suffers from data sparsity issues, especially for new users. One promising line of work is the cross-domain recommendation, which trains models with data across multiple domains to improve the performance in data-scarce domains. Recent proposed cross-domain sequential recommendation models such as PiNet and DASL have a common drawback relying heavily on overlapped users in different domains, which limits their usage in practical recommender systems. In this paper, we propose a M ixed A ttention N etwork (MAN) with local and global attention modules to extract the domain-specific and cross-domain information. Firstly, we propose a local/global encoding layer to capture the domain-specific/cross-domain sequential pattern. Then we propose a mixed attention layer with item similarity attention, sequence-fusion attention, and group-prototype attention to capture the local/global item similarity, fuse the local/global item sequence, and extract the user groups across different domains, respectively. Finally, we propose a local/global prediction layer to further evolve and combine the domain-specific and cross-domain interests. Experimental results on two real-world datasets (each with two domains) demonstrate the superiority of our proposed model. Further study also illustrates that our proposed method and components are model-agnostic and effective, respectively. The code and data are available at https://github.com/Guanyu-Lin/MAN.

Modern personalized recommendation services often rely on user feedback, either explicit or implicit, to improve the quality of services. Explicit feedback refers to behaviors like ratings, while implicit feedback refers to behaviors like user clicks. However, in the scenario of full-screen video viewing experiences like Tiktok and Reels, the click action is absent, resulting in unclear feedback from users, hence introducing noises in modeling training. Existing approaches on de-noising recommendation mainly focus on positive instances while ignoring the noise in a large amount of sampled negative feedback. In this paper, we propose a meta-learning method to annotate the unlabeled data from loss and gradient perspectives, which considers the noises in both positive and negative instances. Specifically, we first propose anInverse Dual Loss (IDL) to boost the true label learning and prevent the false label learning. Then we further propose anInverse Gradient (IG) method to explore the correct updating gradient and adjust the updating based on meta-learning. Finally, we conduct extensive experiments on both benchmark and industrial datasets where our proposed method can significantly improve AUC by 9.25% against state-of-the-art methods. Further analysis verifies the proposed inverse learning framework is model-agnostic and can improve a variety of recommendation backbones. The source code, along with the best hyper-parameter settings, is available at this link: https://github.com/Guanyu-Lin/InverseLearning.

The recommender systems on online platforms assist users in finding personalized information, yet this also leads to the issue of limited diversity, potentially giving rise to societal issues such as filter bubbles. Despite significant progress in diversified recommendation algorithms, they have not been extensively experimented with and evaluated for effectiveness in large-scale, full-stage industrial recommender systems. Specifically, industrial recommenders usually consist of three stages of matching, ranking, and re-ranking, in which specific characteristics lead to critical challenges for promoting both recommendation diversity and user engagement. First, user interests are partially observed due to only relevance maximization. Second, item-side feature-aware bias causes imbalanced recommendations. Last, the impact of diversity perception on user engagement stresses the necessity of explicit diversity modeling. To address these challenges in industrial systems, in this work, we deploy several existing diversified algorithms in a real-world short-video platform, including exploration-exploitation, feature-aware debiasing, and diversity optimization. We conduct large-scale online A/B testing for evaluation via online metrics of user engagement and recommendation diversity. Performance improvement across full stages demonstrates the effectiveness of these simple solutions. From comparing performance across different stages and algorithms, we identify that the ranking stage is the most suitable for real-world deployment, and the combination of debiasing and diversity optimization is a promising direction in terms of diversified recommendations. This work provides experiential guidance for the large-scale deployment of diversified algorithms and the construction of a more inclusive platform on the Web.

Multimodal recommender systems have acquired applications in broad web scenarios such as e-commerce businesses and short-video platforms. Existing multimodal recommendation methods generally boost performance by introducing item-side multimodal content as supplement information. However, the common training paradigm, i.e., encoding unimodal content respectively and fusing them to fit user preference scores, makes the model biased towards items with prevailing modality content under non-uniform training data. This results in a serious item-side unfairness issue, i.e., some items with prevailing modality content are over-recommended while a large number of items don't receive adequate recommendation opportunities, leaving corresponding content providers at great disadvantage. Aiming to eliminate such modality bias and promote item-side fairness, we propose a fairness-aware modality debiasing framework based on counterfactual inference. In the training stage, we additionally introduce unimodal prediction branches to capture the modality bias. In the inference stage, we conduct a fairness-aware counterfactual inference to adaptively eliminate the modality bias. The proposed framework is model-agnostic and flexible to be implemented in various multimodal recommendation models. Extensive experiments on two datasets demonstrate that the proposed method can significantly enhance item-side fairness while providing competitive recommendation accuracy. Our proposed framework is expected to help mitigate the unfair treatment experienced by vulnerable content providers on multimedia web platforms. Codes are available in https://github.com/tsinghua-fib-lab-WWW2024-Modality-Debiasing.

Filter bubbles have been studied extensively within the context of online content platforms due to their potential to cause undesirable outcomes such as user dissatisfaction or polarization. With the rise of short-video platforms, the filter bubble has been given extra attention because these platforms rely on an unprecedented use of the recommender system to provide relevant content. In our work, we investigate the deep filter bubble, which refers to the user being exposed to narrow content within their broad interests. We accomplish this using one-year interaction data from a top short-video platform in China, which includes hierarchical data with three levels of categories for each video. We formalize our definition of a "deep" filter bubble within this context, and then explore various correlations within the data: first understanding the evolution of the deep filter bubble over time, and later revealing some of the factors that give rise to this phenomenon, such as specific categories, user demographics, and feedback type. We observe that while the overall proportion of users in a filter bubble remains largely constant over time, the depth composition of their filter bubble changes. In addition, we find that some demographic groups that have a higher likelihood of seeing narrower content and implicit feedback signals can lead to less bubble formation. Finally, we propose some ways in which recommender systems can be designed to reduce the risk of a user getting caught in a bubble.

Accurately predicting individual-level infection state is of great value since its essential role in reducing the damage of the epidemic. However, there exists an inescapable risk of privacy leakage in the fine-grained user mobility trajectories required by individual-level infection prediction. In this article, we focus on developing a framework of privacy-preserving individual-level infection prediction based on federated learning (FL) and graph neural networks (GNN). We proposeFalcon, aFederated grAphLearning method for privacy-preserving individual-level infeCtion predictiON. It utilizes a novel hypergraph structure with spatio-temporal hyperedges to describe the complex interactions between individuals and locations in the contagion process. By organically combining the FL framework with hypergraph neural networks, the information propagation process of the graph machine learning is able to be divided into two stages distributed on the server and the clients, respectively, so as to effectively protect user privacy while transmitting high-level information. Furthermore, it elaborately designs a differential privacy perturbation mechanism as well as a plausible pseudo location generation approach to preserve user privacy in the graph structure. Besides, it introduces a cooperative coupling mechanism between the individual-level prediction model and an additional region-level model to mitigate the detrimental impacts caused by the injected obfuscation mechanisms. Extensive experimental results show that our methodology outperforms state-of-the-art algorithms and is able to protect user privacy against actual privacy attacks. Our code and datasets are available at the link: https://github.com/wjfu99/FL-epidemic .

Recommender systems have become crucial in information filtering nowadays. Existing recommender systems extract user preferences based on the correlation in data, such as behavioral correlation in collaborative filtering, feature-feature, or feature-behavior correlation in click-through rate prediction. However, unfortunately, the real world is driven by causality , not just correlation, and correlation does not imply causation. For instance, recommender systems might recommend a battery charger to a user after buying a phone, where the latter can serve as the cause of the former; such a causal relation cannot be reversed. Recently, to address this, researchers in recommender systems have begun utilizing causal inference to extract causality, thereby enhancing the recommender system. In this survey, we offer a comprehensive review of the literature on causal inference-based recommendation. Initially, we introduce the fundamental concepts of both recommender system and causal inference as the foundation for subsequent content. We then highlight the typical issues faced by non-causality recommender system. Following that, we thoroughly review the existing work on causal inference-based recommender systems, based on a taxonomy of three-aspect challenges that causal inference can address. Finally, we discuss the open problems in this critical research area and suggest important potential future works.

Multi-types of behaviors (e.g., clicking, carting, purchasing, etc.) widely exist in most real-world recommendation scenarios, which are beneficial to learn users’ multi-faceted preferences. As dependencies are explicitly exhibited by the multiple types of behaviors, effectively modeling complex behavior dependencies is crucial for multi-behavior prediction. The state-of-the-art multi-behavior models learn behavior dependencies indistinguishably with all historical interactions as input. However, different behaviors may reflect different aspects of user preference, which means that some irrelevant interactions may play as noises to the target behavior to be predicted. To address the aforementioned limitations, we introduce multi-interest learning to the multi-behavior recommendation. More specifically, we propose a novel Coarse-to-fine Knowledge-enhanced Multi-interest Learning (CKML) framework to learn shared and behavior-specific interests for different behaviors. CKML introduces two advanced modules, namely Coarse-grained Interest Extracting (CIE) and Fine-grained Behavioral Correlation (FBC) , which work jointly to capture fine-grained behavioral dependencies. CIE uses knowledge-aware information to extract initial representations of each interest. FBC incorporates a dynamic routing scheme to further assign each behavior among interests. Empirical results on three real-world datasets verify the effectiveness and efficiency of our model in exploiting multi-behavior data.

Knowledge graphs (KGs) can help enhance recommendations, especially for the data-sparsity scenarios with limited user-item interaction data. Due to the strong power of representation learning of graph neural networks (GNNs), recent works of KG-based recommendation deploy GNN models to learn from both knowledge graph and user-item bipartite interaction graph. However, these works have not well considered the hierarchical structure of knowledge graph, leading to sub-optimal results. Despite the benefit of hierarchical structure, leveraging it is challenging since the structure is always partly-observed. In this work, we first propose to reveal unknown hierarchical structures with a supervised signal detection method and then exploit the hierarchical structure with disentangling representation learning. We conduct experiments on two large-scale datasets, of which the results well verify the superiority and rationality of the proposed method. Further experiments of ablation study with respect to key model designs have demonstrated the effectiveness and rationality of our proposed model. The code is available at https://github.com/tsinghua-fib-lab/HIKE .

Micro-video platforms such as TikTok are extremely popular nowadays. One important feature is that users no longer select interested videos from a set; instead, they either watch the recommended video or skip to the next one. As a result, the time length of users’ watching behavior becomes the most important signal for identifying preferences. However, our empirical data analysis has shown a video-length effect that long videos can more easily receive a higher value of average view time, and thus adopting such view-time labels for measuring user preferences can easily induce a biased model that favors the longer videos. In this article, we propose a V ideo L ength D ebiasing Rec ommendation (VLDRec) method to alleviate such an effect for micro-video recommendation. VLDRec designs the data labeling approach and the sample generation module that better capture user preferences in a view-time-oriented manner. It further leverages the multi-task learning technique to jointly optimize the above samples with the original biased ones. Extensive experiments show that VLDRec can improve users’ view time by 1.81% and 11.32% on two real-world datasets, given a recommendation list of a fixed overall video length, compared with the best baseline method. Moreover, VLDRec is also more effective in matching users’ interests in terms of the video content.

Multi-behavior recommendation exploits multiple types of user-item interactions, such as view and cart , to learn user preferences and has demonstrated to be an effective solution to alleviate the data sparsity problem faced by the traditional models that often utilize only one type of interaction for recommendation. In real scenarios, users often take a sequence of actions to interact with an item, in order to get more information about the item and thus accurately evaluate whether an item fits their personal preferences. Those interaction behaviors often obey a certain order, and more importantly, different behaviors reveal different information or aspects of user preferences towards the target item. Most existing multi-behavior recommendation methods take the strategy to first extract information from different behaviors separately and then fuse them for final prediction. However, they have not exploited the connections between different behaviors to learn user preferences. Besides, they often introduce complex model structures and more parameters to model multiple behaviors, largely increasing the space and time complexity. In this work, we propose a lightweight multi-behavior recommendation model named Cascading Residual Graph Convolutional Network ( CRGCN for short) for multi-behavior recommendation, which can explicitly exploit the connections between different behaviors into the embedding learning process without introducing any additional parameters (with comparison to the single-behavior based recommendation model). In particular, we design a cascading residual graph convolutional network (GCN) structure, which enables our model to learn user preferences by continuously refining the embeddings across different types of behaviors. The multi-task learning method is adopted to jointly optimize our model based on different behaviors. Extensive experimental results on three real-world benchmark datasets show that CRGCN can substantially outperform the state-of-the-art methods, achieving 24.76%, 27.28%, and 25.10% relative gains on average in terms of HR@K (K = {10,20,50,80}) over the best baseline across the three datasets. Further studies also analyze the effects of leveraging multi-behaviors in different numbers and orders on the final performance.

Short-video recommendation is one of the most important recommendation applications in today's industrial information systems. Compared with other recommendation tasks, the enormous amount of feedback is the most typical characteristic. Specifically, in short-video recommendation, the easiest-to-collect user feedback is theskipping behavior, which leads to two critical challenges for the recommendation model. First, the skipping behavior reflects implicit user preferences, and thus, it is challenging for interest extraction. Second, this kind of special feedback involves multiple objectives, such as total watching time and skipping rate, which is also very challenging. In this paper, we present our industrial solution in Kuaishou1, which serves billion-level users every day. Specifically, we deploy a feedback-aware encoding module that extracts user preferences, taking the impact of context into consideration. We further design a multi-objective prediction module which well distinguishes the relation and differences among different model objectives in the short-video recommendation. We conduct extensive online A/B tests, along with detailed and careful analysis, which verify the effectiveness of our solution.

Location prediction acts as a fundamental service in today's location-based information platform, which helps users access locations satisfying their demands, improving both user experience and platform profit. Since users with unambiguous demands prefer specific locations while users with compound demands consider first regions and then specific locations, it is necessary to model multi-grained user preferences at different geographical scales. However, most of the existing works concentrate on user preferences at the location-scale only, which can not understand users traveling behaviors thoroughly. In this paper, we propose to model both the fine-grained user preferences at the location scale and the coarsegrained user preferences at the region scale. Specifically, the proposed model harnesses the efficient information extraction power of graph neural networks. Moreover, the proposed geographical calibration method also helps to capture multi-grained user preferences accurately. Experiments on datasets of two very large cities demonstrate the significant performance improvement using our approach over state-of-the-art models. We also conduct experiments to further demonstrate the effectiveness of each component in the proposed model. Source codes of this paper are available at https://github.com/tsinghua-fib-lab/SIGSPATIAL-MMGUP/.

Living needs refer to the various needs in human's daily lives for survival and well-being, including food, housing, entertainment, etc. At life service platforms that connect users to service providers, such as Meituan, the problem of living needs prediction is fundamental as it helps understand users and boost various downstream applications such as personalized recommendation. However, the problem has not been well explored and is faced with two critical challenges. First, the needs are naturally connected to specific locations and times, suffering from complex impacts from the spatiotemporal context. Second, there is a significant gap between users' actual living needs and their historical records on the platform. To address these two challenges, we design a system of living NEeds predictiON named NEON, consisting of three phases: feature mining, feature fusion and multi-task prediction. In the feature mining phase, we carefully extract individual-level user features for spatiotemporal modeling, and aggregated-level behavioral features for enriching data, which serve as the basis for addressing two challenges, respectively. Further, in the feature fusion phase, we propose a neural network that effectively fuses two parts of features into the user representation. Moreover, we design a multitask prediction phase, where the auxiliary task of needs-meeting way prediction can enhance the modeling of spatiotemporal context. Extensive offline evaluations verify that our NEON system can effectively predict users' living needs. Furthermore, we deploy NEON into Meituan's algorithm engine and evaluate how it enhances the three downstream prediction applications, via large-scale online A/B testing. As a representative result, deploying our system leads to a 1.886% increase w.r.t. CTCVR in Meituan homepage recommendation. The results demonstrate NEON's effectiveness in predicting fine-grained user needs, needs-meeting way, and potential needs, highlighting the immense application value of NEON.

Understanding and characterizing the vulnerability of urban infrastructures, which refers to the engineering facilities essential for the regular running of cities and that exist naturally in the form of networks, is of great value to us. Potential applications include protecting fragile facilities and designing robust topologies, etc. Due to the strong correlation between different topological characteristics and infrastructure vulnerability and their complicated evolution mechanisms, some heuristic and machine assisted analysis fall short in addressing such a scenario. In this paper, we model the interdependent network as a heterogeneous graph and propose a system based on graph neural network with reinforcement learning, which can be trained on real-world data, to characterize the vulnerability of the city system accurately. The presented system leverages deep learning techniques to understand and analyze the heterogeneous graph, which enables us to capture the risk of cascade failure and discover vulnerable infrastructures of cities. Extensive experiments with various requests demonstrate not only the expressive power of our system but also transferring ability and necessity of the specific components. All source codes and models including those that can reproduce all figures analyzed in this work are publicly available at this link: https://github.com/tsinghua-fib-lab/KDD2023-ID546-UrbanInfra.

Automated Machine Learning (AutoML) techniques have recently been introduced to design Collaborative Filtering (CF) models in a data-specific manner. However, existing works either search architectures or hyperparameters while ignoring the fact they are intrinsically related and should be considered together. This motivates us to consider a joint hyperparameter and architecture search method to design CF models. However, this is not easy because of the large search space and high evaluation cost. To solve these challenges, we reduce the space by screening out usefulness hyperparameter choices through a comprehensive understanding of individual hyperparameters. Next, we propose a two-stage search algorithm to find proper configurations from the reduced space. In the first stage, we leverage knowledge from subsampled datasets to reduce evaluation costs; in the second stage, we efficiently fine-tune top candidate models on the whole dataset. Extensive experiments on real-world datasets show better performance can be achieved compared with both hand-designed and previous searched models. Besides, ablation and case studies demonstrate the effectiveness of our search framework.

Sequential recommendation is one of the most important tasks in recommender systems, which aims to recommend the next interacted item with historical behaviors as input. Traditional sequential recommendation always mainly considers the collected positive feedback such as click, purchase, etc. However, in short-video platforms such as TikTok, video viewing behavior may not always represent positive feedback. Specifically, the videos are played automatically, and users passively receive the recommended videos. In this new scenario, users passively express negative feedback by skipping over videos they do not like, which provides valuable information about their preferences. Different from the negative feedback studied in traditional recommender systems, this passive-negative feedback can reflect users’ interests and serve as an important supervision signal in extracting users’ preferences. Therefore, it is essential to carefully design and utilize it in this novel recommendation scenario. In this work, we first conduct analyses based on a large-scale real-world short-video behavior dataset and illustrate the significance of leveraging passive feedback. We then propose a novel method that deploys the sub-interest encoder, which incorporates positive feedback and passive-negative feedback as supervision signals to learn the user’s current active sub-interest. Moreover, we introduce an adaptive fusion layer to integrate various sub-interests effectively. To enhance the robustness of our model, we then introduce a multi-task learning module to simultaneously optimize two kinds of feedback – passive-negative feedback and traditional randomly-sampled negative feedback. The experiments on two large-scale datasets verify that the proposed method can significantly outperform state-of-the-art approaches. The code is released at https://github.com/tsinghua-fib-lab/RecSys2023-SINE to benefit the community.

Graph Neural Network (GNN)-based models have become the mainstream approach for recommender systems. Despite the effectiveness, they are still suffering from the cold-start problem, i.e., recommend for few-interaction items. Existing GNN-based recommendation models to address the cold-start problem mainly focus on utilizing auxiliary features of users and items, leaving the user-item interactions under-utilized. However, embeddings distributions of cold and warm items are still largely different, since cold items' embeddings are learned from lower-popularity interactions, while warm items' embeddings are from higher-popularity interactions. Thus, there is a seesaw phenomenon, where the recommendation performance for the cold and warm items cannot be improved simultaneously. To this end, we proposed a Uncertainty-aware Consistency learning framework for Cold-start item recommendation (shorten as UCC) solely based on user-item interactions. Under this framework, we train the teacher model (generator) and student model (recommender) with consistency learning, to ensure the cold items with additionally generated low-uncertainty interactions can have similar distribution with the warm items. Therefore, the proposed framework improves the recommendation of cold and warm items at the same time, without hurting any one of them. Extensive experiments on benchmark datasets demonstrate that our proposed method significantly outperforms state-of-the-art methods on both warm and cold items, with an average performance improvement of 27.6%.

Recent years have witnessed the rapid development of online micro-video platforms, in which the recommender system plays an essential role in overcoming the information overloading problem and providing personalized content for users. Although some progress has been achieved in the micro-video recommendation, there are still some limitations in learning the representations of user interests and video features. Specifically, the user modeling in existing works is performed at a coarse-grained level, i.e., video level. However, in micro-video recommendation, the user feedback is at a continuous form---users can skip over a video at each frame---which reveals fine-grained user preferences. In this work, we approach the problem of learning fine-grained user preferences for micro-video recommendation by first collecting two real-world datasets. To address the challenges of preference modeling and weak supervision signal, we propose a solution named FRAME (short for Fine-gRAined preference-modeling for Micro-video rEcommendation). Specifically, we first adopt visual feature extraction and transformation to maintain the fine-grained video embeddings. We then propose graph convolution layers to learn the user preference from complex and fine-grained user-clip relations, and hybrid-supervision objectives for enhancing the supervision signal. The experimental results on two collected real-world datasets demonstrate the effectiveness of our proposed model. We release the datasets and codes in https://github.com/tsinghua-fib-lab/FRAME, which we believe can benefit the community.

In the information-overloaded era of the Web, recommender systems that provide personalized content filtering are now the mainstream portal for users to access Web information. Recommender systems deploy machine learning models to learn users’ preferences from collected historical data, leading to more centralized recommendation results due to the feedback loop. As a result, it will harm the ranking of content outside the narrowed scope and limit the options seen by users. In this work, we first conduct data analysis from a graph view to observe that the users’ feedback is restricted to limited items, verifying the phenomenon of centralized recommendation. We further develop a general simulation framework to derive the procedure of the recommender system, including data collection, model learning, and item exposure, which forms a loop. To address the filter bubble issue under the feedback loop, we then propose a general and easy-to-use reinforcement learning-based method, which can adaptively select few but effective connections between nodes from different communities as the exposure list. We conduct extensive experiments in the simulation framework based on large-scale real-world datasets. The results demonstrate that our proposed reinforcement learning-based control method can serve as an effective solution to alleviate the filter bubble and the separated communities induced by it. We believe the proposed framework of controllable recommendation in this work can inspire not only the researchers of recommender systems, but also a broader community concerned with artificial intelligence algorithms’ impact on humanity, especially for those vulnerable populations on the Web.

Accurate user interest modeling is vital for recommendation scenarios. One of the effective solutions is the sequential recommendation that relies on click behaviors, but this is not elegant in the video feed recommendation where users are passive in receiving the streaming contents and return skip or no-skip behaviors. Here skip and no-skip behaviors can be treated as negative and positive feedback, respectively. With the mixture of positive and negative feedback, it is challenging to capture the transition pattern of behavioral sequence. To do so, FeedRec has exploited a shared vanilla Transformer, which may be inelegant because head interaction of multi-heads attention does not consider different types of feedback. In this paper, we propose Dual-interest Factorization-heads Attention for Sequential Recommendation (short for DFAR) consisting of feedback-aware encoding layer, dual-interest disentangling layer and prediction layer. In the feedback-aware encoding layer, we first suppose each head of multi-heads attention can capture specific feedback relations. Then we further propose factorization-heads attention which can mask specific head interaction and inject feedback information so as to factorize the relation between different types of feedback. Additionally, we propose a dual-interest disentangling layer to decouple positive and negative interests before performing disentanglement on their representations. Finally, we evolve the positive and negative interests by corresponding towers whose outputs are contrastive by BPR loss. Experiments on two real-world datasets show the superiority of our proposed method against state-of-the-art baselines. Further ablation study and visualization also sustain its effectiveness. We release the source code here: https://github.com/tsinghua-fib-lab/WWW2023-DFAR.

Graph Neural Network (GNN) based social recommendation models improve the prediction accuracy of user preference by leveraging GNN in exploiting preference similarity contained in social relations. However, in terms of both effectiveness and efficiency of recommendation, a large portion of social relations can be redundant or even noisy, e.g., it is quite normal that friends share no preference in a certain domain. Existing models do not fully solve this problem of relation redundancy and noise, as they directly characterize social influence over the full social network. In this paper, we instead propose to improve graph based social recommendation by only retaining the informative social relations to ensure an efficient and effective influence diffusion, i.e., graph denoising. Our designed denoising method is preference-guided to model social relation confidence and benefits user preference learning in return by providing a denoised but more informative social graph for recommendation models. Moreover, to avoid interference of noisy social relations, it designs a self-correcting curriculum learning module and an adaptive denoising strategy, both favoring highly-confident samples. Experimental results on three public datasets demonstrate its consistent capability of improving three state-of-the-art social recommendation models by robustly removing 10-40% of original relations. We release the source code at https://github.com/tsinghua-fib-lab/Graph-Denoising-SocialRec.

Bundle recommendation aims to recommend a bundle of items for a user to consume as a whole. Related work can be divided into two categories: 1) to recommend the platforms prebuilt bundles to users; 2) generate personalized bundles for users. These two problems are not well solved. In this work, we propose two graph neural network models, a BGCN model for prebuilt bundle recommendation, and a BGGN model for personalized bundle generation. First, BGCN unifies the user-item interaction, the user-bundle interaction and the bundle-item affiliation into a heterogeneous graph. With item nodes as the bridge, graph convolutional propagation between user and bundle nodes makes the learned representations capture the item-level semantics. Second, BGGN re-constructs bundles into graphs based on the item co-occurrence pattern and the users supervision signal. The complex and high-order item-item relationships in the bundle graph are explicitly modeled through graph generation. Empirical results demonstrate the substantial performance gains of BGCN and BGGN. We have released the datasets and codes at this link: https://github.com/cjx0525/BGCN.

Social e-commerce uses social media as a new prevalent platform for online shopping. In this paper, we address the problem of cross-platform recommendation for social e-commerce, i.e., recommending products to users when they are shopping through social media. To the best of our knowledge, this is a new and important problem for all e-commerce companies (e.g. Amazon, Alibaba), but has never been studied before. Existing cross-platform and social related recommendation methods cannot be applied directly to this problem since they do not co-consider the social information and the cross-platform characteristics together. To study this problem, we collect two real-world datasets from social e-commerce services. We first investigate the heterogeneous shopping behaviors between traditional e-commerce app and social media. Based on these observations from data, we propose CROSS (Cross-platform Recommendation for Online Shopping in Social Media), a recommendation framework utilizing not only user-item interaction data on both platforms, but also social relation data on social media. The framework is general and we propose two variants, CROSS-MF and CROSS-NCF. Extensive experiments on two real-world social e-commerce datasets demonstrate that our proposed CROSS significantly outperforms existing state-of-the-art methods.

Social recommendation leverages social information to alleviate data sparsity and cold-start issues of collaborative filtering (CF) methods. Most existing works model user interests following the assumption ofsocial homophilybased on social-relation data. The explicit modeling ofsocial influence, which also largely affects user behaviors, has not been well explored. Considering user behaviors may be driven by social factors in today’s information services (e.g., purchasing products shared by close friends on social e-commerce applications), these methods will be suboptimal. In this work, we propose a method modeling both social homophily-aware user interests and social influence as two essential effects on user behaviors for social recommendation, named as DISGCN (short forDISentangled modeling of Social homophily and influence withGraphConvolutionalNetwork). Specifically, we devise a disentangled embedding layer to encode these two effects. Furthermore, two tailored graph convolutional layers are developed to disentangle them refinedly, leveraging the high-order embedding propagation in social-network graph from two aspects. Technically, first, the operation of attentive embedding propagation is adopted for capturing personalized social homophily-aware interests, and second, the item-gate-based embedding propagation is proposed for capturing item-specific social influence. In addition, to ensure the disentanglement of social influence, we propose a contrastive learning framework that endows corresponding embeddings with explicit semantics. Extensive experiments on two real-world datasets demonstrate the effectiveness of our proposed model. Further studies also verify the rationality and necessity of our designs. We have released the datasets and codes at this link:https://github.com/tsinghua-fib-lab/DISGCN.

Point-of-Interest (POI) recommendation has drawn a lot of attention in both academia and industry. It utilizes user check-in data, aiming at recommending unvisited POIs to users. To address the data-sparsity problem, geographical information of POIs is often incorporated into recommender systems. However, most of the existing approaches model geographical impact in an implicit way, in which geographical information is encoded as auxiliary vectors for learning unified representations of users and POIs. Following this paradigm, the embedding of POIs can not reflect geographical similarity directly; thus, an explicit modeling approach is needed as geography is of great importance in POI recommendation. To address challenges in disentangling geographical effect, we proposed a disentangled representation learning method named DIG (short for Disentangled embedding of user Interest and POIs' Geographical information). Aiming at decoupling the geographical factor and the user interest factor thoroughly, we first proposed a geo-constrained negative sampling strategy, which helps to find reliable negative samples for the two factors. Second, a geo-enhanced soft-weighted loss function was proposed to quantify the trade-off between the two factors in loss computation. Extensive experiments have been conducted on two real-world datasets, and results have demonstrated the significant improvement of DIG at 3.92% - 20.32% 3.92% - 20.32% on recall, and 2.53% - 11.48% 2.53% - 11.48% on hit ratio, compared with other state-of-the-art approaches.

In this work, we aim at developing an effective method to predict user purchase intention with the focus on the price factor in recommender systems. The main difficulties are two-fold: 1) the preference and sensitivity of a user on item price are unknown, which are only implicitly reflected in the items that the user has purchased, and 2) how the item price affects a users intention depends largely on the product category, that is, the perception and affordability of a user on item price could vary significantly across categories. Towards the first difficulty, we propose to model the transitive relationship between user-to-item and item-to-price, taking the inspiration from the recently developed Graph Convolution Networks (GCN). The key idea is to propagate the influence of price on users with items as the bridge, so as to make the learned user representations be price-aware. For the second difficulty, we further integrate item categories into the propagation progress and model the possible pairwise interactions for predicting user-item interactions. We conduct extensive experiments on two real-world datasets, demonstrating the effectiveness of our GCN-based method in learning the price-aware preference of users.

Recommender system is one of the most important information services on today’s Internet. Recently, graph neural networks have become the new state-of-the-art approach to recommender systems. In this survey, we conduct a comprehensive review of the literature on graph neural network-based recommender systems. We first introduce the background and the history of the development of both recommender systems and graph neural networks. For recommender systems, in general, there are four aspects for categorizing existing works: stage, scenario, objective, and application. For graph neural networks, the existing methods consist of two categories: spectral models and spatial ones. We then discuss the motivation of applying graph neural networks into recommender systems, mainly consisting of the high-order connectivity, the structural property of data and the enhanced supervision signal. We then systematically analyze the challenges in graph construction, embedding propagation/aggregation, model optimization, and computation efficiency. Afterward and primarily, we provide a comprehensive overview of a multitude of existing works of graph neural network-based recommender systems, following the taxonomy above. Finally, we raise discussions on the open problems and promising future directions in this area. We summarize the representative papers along with their code repositories in https://github.com/tsinghua-fib-lab/GNN-Recommender-Systems .

Session-based recommendation (SBR) aims to recommend items based on user behaviors in a session. For the online life service platforms, such as Meituan, both the user's location and the current time primarily cause the different patterns and intents in user behaviors. Hence, spatiotemporal context plays a significant role in the recommendation on those platforms, which motivates an important problem of spatiotemporal-aware session-based recommendation (STSBR). Since the spatiotemporal context is introduced, there are two critical challenges: 1) how to capture session-level relations of spatiotemporal context (inter-session view), and 2) how to model the complex user decision-making process at a specific location and time (intra-session view). To address them, we propose a novel solution named STAGE in this paper. Specifically, STAGE first constructs a global information graph to model the multi-level relations among all sessions, and a session decision graph to capture the complex user decision process for each session. STAGE then performs inter-session and intra-session embedding propagation on the constructed graphs with the proposed graph attentive convolution (GAC) to learn representations from the above two perspectives. Finally, the learned representations are combined with spatiotemporal-aware soft-attention for final recommendation. Extensive experiments on two datasets from Meituan demonstrate the superiority of STAGE over state-of-the-art methods. Further studies also verify that each component is effective.

In recent years, short-form video platforms have emerged rapidly and attracted a large and wide variety of users, with the help of advanced recommendation algorithms. Despite the great success, the algorithms have caused some negative effects, such as information cocoon, algorithm unfairness,etc. In this work, we focus on theinformation cocoon that measures overwhelmingly homogeneity of users' video consumption. Specifically, we conduct an exploratory study of this phenomenon on a top short-form video platform, with one-year behavioral records of new users. First, we evaluate the evolution of users' information cocoons and find the limitation of the diversity of video content that users consume. In addition, we further explore user cocoons via the correlation analysis from three aspects, including user demographics, video content, and user-recommender interactions driven by algorithms and user preferences. Correspondingly, we observe that video content plays a more significant role in affecting user cocoons than demographics does. In terms of user-recommender interactions, more accurate personalization does not contribute to more severe information cocoons necessarily, while users with narrow preferences are more likely to be trapped. In summary, our study illuminates the current concern of information cocoons that may hurt user experience on short-form video platforms, and offers potential directions for mitigation implied by the correlation analysis.

Consumption intent, defined as the decision-driven force of consumption behaviors, is crucial for improving the explainability and performance of user-modeling systems, with various downstream applications like recommendation and targeted marketing. However, consumption intent is implicit, and only a few known intents have been explored from the user consumption data in Meituan. Hence, discovering new consumption intents is a crucial but challenging task, which suffers from two critical challenges: 1) how to encode the consumption intent related to multiple aspects of preferences, and 2) how to discover the new intents with only a few known ones. In Meituan, we designed the AutoIntent system, consisting of the disentangled intent encoder and intent discovery decoder, to address the above challenges. Specifically, for the disentangled intent encoder, we construct three groups of dual hypergraphs to capture the high-order relations under the three aspects of preferences and then utilize the designed hypergraph neural networks to extract disentangled intent features. For the intent discovery decoder, we propose to build intent-pair pseudo labels based on the denoised feature similarities to transfer knowledge from known intents to new ones. Extensive offline evaluations verify that AutoIntent can effectively discover unknown consumption intents. Moreover, we deploy AutoIntent in the recommendation engine of the Meituan APP, and the further online evaluation verifies its effectiveness.

In online information systems, users make decisions based on factors of several specific aspects, such as brand, price, etc. Existing recommendation engines ignore the explicit modeling of these factors, leading to sub-optimal recommendation performance. In this paper, we focus on the real-world scenario where these factors can be explicitly captured (the users are exposed with decision factor-based persuasion texts, i.e., persuasion factors). Although it allows us for explicit modeling of user-decision process, there are critical challenges including the persuasion factor's representation learning and effect estimation, along with the data-sparsity problem. To address them, in this work, we present our POEM (short for Persuasion factOr Effect Modeling) system. We first propose the persuasion-factor graph convolutional layers for encoding and learning representations from the persuasion-aware interaction data. Then we develop a prediction layer that fully considers the user sensitivity to the persuasion factors. Finally, to address the data-sparsity issue, we propose a counterfactual learning-based data augmentation method to enhance the supervision signal. Real-world experiments demonstrate the effectiveness of our proposed framework of modeling the effect of persuasion factors.

Session-based recommendation (SBR) aims at the next-item prediction with a short behavior session. Existing solutions fail to address two main challenges: 1) user interests are shown as dynamically coupled intents, and 2) sessions always contain noisy signals. To address them, in this paper, we propose a hypergraph-based solution, HIDE. Specifically, HIDE first constructs a hypergraph for each session to model the possible interest transitions from distinct perspectives. HIDE then disentangles the intents under each item click in micro and macro manners. In the micro-disentanglement, we perform intent-aware embedding propagation on session hypergraph to adaptively activate disentangled intents from noisy data. In the macro-disentanglement, we introduce an auxiliary intent-classification task to encourage the independence of different intents. Finally, we generate the intent-specific representations for the given session to make the final recommendation. Benchmark evaluations demonstrate the significant performance gain of our HIDE over the state-of-the-art methods.

Widely applied in today's recommender systems, sequential recommendation predicts the next interacted item for a given user via his/her historical item sequence. However, sequential recommendation suffers data sparsity issue like most recommenders. To extract auxiliary signals from the data, some recent works exploit self-supervised learning to generate augmented data via dropout strategy, which, however, leads to sparser sequential data and obscure signals. In this paper, we propose D ual C ontrastive N etwork (DCN) to boost sequential recommendation, from a new perspective of integrating auxiliary user-sequence for items. Specifically, we propose two kinds of contrastive learning. The first one is the dual representation contrastive learning that minimizes the distances between embeddings and sequence-representations of users/items. The second one is the dual interest contrastive learning which aims to self-supervise the static interest with the dynamic interest of next item prediction via auxiliary training. We also incorporate the auxiliary task of predicting next user for a given item's historical user sequence, which can capture the trends of items preferred by certain types of users. Experiments on benchmark datasets verify the effectiveness of our proposed method. Further ablation study also illustrates the boosting effect of the proposed components upon different sequential models.

Recently, graph neural network (GNN) has become the new state-of-the-art approach in many recommendation problems, with its strong ability to handle structured data and to explore high-order information. However, as the recommendation tasks are diverse and various in the real world, it is quite challenging to design proper GNN methods for specific problems. In this tutorial, we focus on the critical challenges of GNN-based recommendation and the potential solutions. Specifically, we start from an extensive background of recommender systems and graph neural networks. Then we fully discuss why GNNs are required in recommender systems and the four parts of challenges, including graph construction, network design, optimization, and computation efficiency. Then, we discuss how to address these challenges by elaborating on the recent advances of GNN-based recommendation models, with a systematic taxonomy from four critical perspectives: stages, scenarios, objectives, and applications. Last, we finalize this tutorial with conclusions and discuss important future directions.

Modeling user’s long-term and short-term interests is crucial for accurate recommendation. However, since there is no manually annotated label for user interests, existing approaches always follow the paradigm of entangling these two aspects, which may lead to inferior recommendation accuracy and interpretability. In this paper, to address it, we propose a Contrastive learning framework to disentangle Long and Short-term interests for Recommendation (CLSR) with self-supervision. Specifically, we first propose two separate encoders to independently capture user interests of different time scales. We then extract long-term and short-term interests proxies from the interaction sequences, which serve as pseudo labels for user interests. Then pairwise contrastive tasks are designed to supervise the similarity between interest representations and their corresponding interest proxies. Finally, since the importance of long-term and short-term interests is dynamically changing, we propose to adaptively aggregate them through an attention-based network for prediction. We conduct experiments on two large-scale real-world datasets for e-commerce and short-video recommendation. Empirical results show that our CLSR consistently outperforms all state-of-the-art models with significant improvements: GAUC is improved by over 0.01, and NDCG is improved by over 4%. Further counterfactual evaluations demonstrate that stronger disentanglement of long and short-term interests is successfully achieved by CLSR. The code and data are available at https://github.com/tsinghua-fib-lab/CLSR.

Web systems that provide the same functionality usually share a certain amount of items. This makes it possible to combine data from different websites to improve recommendation quality, known as the cross-domain recommendation task. Despite many research efforts on this task, the main drawback is that they largely assume the data of different systems can be fully shared . Such an assumption is unrealistic different systems are typically operated by different companies, and it may violate business privacy policy to directly share user behavior data since it is highly sensitive. In this work, we consider a more practical scenario to perform cross-domain recommendation. To avoid the leak of user privacy during the data sharing process, we consider sharing only the information of the item side, rather than user behavior data. Specifically, we transfer the item embeddings across domains, making it easier for two companies to reach a consensus (e.g., legal policy) on data sharing since the data to be shared is user-irrelevant and has no explicit semantics. To distill useful signals from transferred item embeddings, we rely on the strong representation power of neural networks and develop a new method named as NATR (short for N eural A ttentive T ransfer R ecommendation ). We perform extensive experiments on two real-world datasets, demonstrating that NATR achieves similar or even better performance than traditional cross-domain recommendation methods that directly share user-relevant data. Further insights are provided on the efficacy of NATR in using the transferred item embeddings to alleviate the data sparsity issue.

Social commerce, which is different from traditional e-commerce where people purchase products via initiative searching or recommendations from the platform, transforms a social community into an inclusive place to do business by enabling people to share products with their friends. A user (sharer), can share a link of a product to their social-connected friends (receiver). Once a receiver purchases the product, the sharer can earn commission provided by the platform. To promote sales, the platform can also assist sharers by providing product candidates which are more likely to be purchased during the social sharing. We define this task of generating sharing suggestions as item recommendation for word-of-mouth scenario, and to the best of our knowledge, this is a new task that has never been explored. In this article, we propose aTriM(short forTriad based word-of-Mouth recommendation) model that can capture both the sharer’s influence and the receiver’s interest at the same time, which are two significant factors that determine whether the receiver will buy the product or not. Furthermore, with joint learning on two parts of interaction data to address data sparsity issue, our proposed TriM-Joint further improves the recommendation performance. By conducting experiments, we show that our proposed models achieve the best results compared to state-of-the-art models with significant improvements by at least$7.4\% \sim 14.4\%$respectively.

Social recommendation, which utilizes social relations to enhance recommender systems, has been gaining increasing attention recently with the rapid development of online social networks. Existing social recommendation methods are based on the assumption, so-calledsocial-trust, that users’ preference or decision is influenced by their social-connected friends’ purchase behaviors. However, they assume that the influences of social relationships are always the same, which violates the fact that users are likely to share preference on different products with different friends. More precisely, friends’ behaviors do not necessarily affect a user’s preferences, and the influence is diverse among different items. In this paper, we contribute a new solution, CSR (short forCharacterizedSocialRegularization) model by designing a universal regularization term for modeling variable social influence. This regularization term captures the finely grained similarity of social-connected friends. We further introduce two variants of our model with different optimization manners. Our proposed model can be applied to both explicit and implicit interaction due to its high generality. Extensive experiments on three real-world datasets demonstrate that our CSR can outperform state-of-the-art social recommendation methods. Further experiments show that CSR can improve recommendation performance for those users with sparse social relations or behavioral interactions.

Group buying, as an emerging form of purchase in social e-commerce websites, such as Pinduoduo1, has recently achieved great success. In this new business model, users, initiator, can launch a group and share products to their social networks, and when there are enough friends, participants, join it, the deal is clinched. Group-buying recommendation for social e-commerce, which recommends an item list when users want to launch a group, plays an important role in the group success ratio and sales. However, designing a personalized recommendation model for group buying is an entirely new problem that is seldom explored. In this work, we take the first step to approach the problem of group-buying recommendation for social e-commerce and develop a GBGCN method (short for Group-Buying Graph Convolutional Network). Considering there are multiple types of behaviors (launch and join) and structured social network data, we first propose to construct directed heterogeneous graphs to represent behavioral data and social networks. We then develop a graph convolutional network model with multi-view embedding propagation, which can extract the complicated high-order graph structure to learn the embeddings. Last, since a failed group-buying implies rich preferences of the initiator and participants, we design a double-pairwise loss function to distill such preference signals. We collect a real-world dataset of group-buying and conduct experiments to evaluate the performance. Empirical results demonstrate that our proposed GBGCN can significantly outperform baseline methods by 2.69%-7.36%. The codes and the dataset are released at https://github.com/Sweetnow/group-buying-recommendation.

Collaborative filtering (CF), as a fundamental approach for recommender systems, is usually built on the latent factor model with learnable parameters to predict users' preferences towards items. However, designing a proper CF model for a given data is not easy, since the properties of datasets are highly diverse. In this paper, motivated by the recent advances in automated machine learning (AutoML), we propose to design a data-specific CF model by AutoML techniques. The key here is a new framework that unifies state-of-the-art (SOTA) CF methods and splits them into disjoint stages of input encoding, embedding function, interaction function, and prediction function. We further develop an easy-to-use, robust, and efficient search strategy, which utilizes random search and a performance predictor for efficient searching within the above framework. In this way, we can combinatorially generalize data-specific CF models, which have not been visited in the literature, from SOTA ones. Extensive experiments on five real-world datasets demonstrate that our method can consistently outperform SOTA ones for various CF tasks. Further experiments verify the rationality of the proposed framework and the efficiency of the search strategy. The searched CF models can also provide insights for exploring more effective methods in the future.

For online life service platforms, such as Meituan, user consumption intention, as the internal driving force of consumption behaviors, plays a significant role in understanding and predicting users' demand and purchase. However, user consumption intention prediction is quite challenging. Different from consumption behaviors, consumption intention is implicit and always not reflected by behavioral data. Moreover, it is affected by both user intrinsic preference and spatio-temporal context. To overcome these challenges, in Meituan, we design a real-world system consisting of two stages, intention detection and prediction. Specifically, at the intention-detection stage, we combine the knowledge of human experts and consumption information to obtain explicit intentions and match consumption with intentions based on user review data. At the intention-prediction stage, to collectively exploit the rich heterogeneous influencing factors, we design a graph neural network-based intention prediction model GRIP, which can capture user intrinsic preference and spatio-temporal context. Extensive offline evaluations demonstrate that our prediction model outperforms the best baseline by 10.26% and 33.28% for two metrics and online A/B tests on millions of users validate the effectiveness of our system.

Sequential recommendation aims to leverage users' historical behaviors to predict their next interaction. Existing works have not yet addressed two main challenges in sequential recommendation. First, user behaviors in their rich historical sequences are often implicit and noisy preference signals, they cannot sufficiently reflect users' actual preferences. In addition, users' dynamic preferences often change rapidly over time, and hence it is difficult to capture user patterns in their historical sequences. In this work, we propose a graph neural network model called SURGE (short forSeqUential Recommendation with Graph neural nEtworks) to address these two issues. Specifically, SURGE integrates different types of preferences in long-term user behaviors into clusters in the graph by re-constructing loose item sequences into tight item-item interest graphs based on metric learning. This helps explicitly distinguish users' core interests, by forming dense clusters in the interest graph. Then, we perform cluster-aware and query-aware graph convolutional propagation and graph pooling on the constructed graph. It dynamically fuses and extracts users' current activated core interests from noisy user behavior sequences. We conduct extensive experiments on both public and proprietary industrial datasets. Experimental results demonstrate significant performance gains of our proposed method compared to state-of-the-art methods. Further studies on sequence length confirm that our method can model long behavioral sequences effectively and efficiently.

These years much effort has been devoted to improving the accuracy or relevance of the recommendation system. Diversity, a crucial factor which measures the dissimilarity among the recommended items, received rather little scrutiny. Directly related to user satisfaction, diversification is usually taken into consideration after generating the candidate items. However, this decoupled design of diversification and candidate generation makes the whole system suboptimal. In this paper, we aim at pushing the diversification to the upstream candidate generation stage, with the help of Graph Convolutional Networks (GCN). Although GCN based recommendation algorithms have shown great power in modeling complex collaborative filtering effect to improve the accuracy of recommendation, how diversity changes is ignored in those advanced works. We propose to perform rebalanced neighbor discovering, category-boosted negative sampling and adversarial learning on top of GCN. We conduct extensive experiments on real-world datasets. Experimental results verify the effectiveness of our proposed method on diversification. Further ablation studies validate that our proposed method significantly alleviates the accuracy-diversity dilemma.

Recommendation models are usually trained on observational interaction data. However, observational interaction data could result from users’ conformity towards popular items, which entangles users’ real interest. Existing methods tracks this problem as eliminating popularity bias, e.g., by re-weighting training samples or leveraging a small fraction of unbiased data. However, the variety of user conformity is ignored by these approaches, and different causes of an interaction are bundled together as unified representations, hence robustness and interpretability are not guaranteed when underlying causes are changing. In this paper, we present DICE, a general framework that learns representations where interest and conformity are structurally disentangled, and various backbone recommendation models could be smoothly integrated. We assign users and items with separate embeddings for interest and conformity, and make each embedding capture only one cause by training with cause-specific data which is obtained according to the colliding effect of causal inference. Our proposed methodology outperforms state-of-the-art baselines with remarkable improvements on two real-world datasets on top of various backbone models. We further demonstrate that the learned embeddings successfully capture the desired causes, and show that DICE guarantees the robustness and interpretability of recommendation.

Most existing recommender systems leverage user behavior data of one type only, such as the purchase behavior in E-commerce that is directly related to the business Key Performance Indicator (KPI) of conversion rate. Besides the key behavioral data, we argue that other forms of user behaviors also provide valuable signal, such as views, clicks, adding a product to shopping carts and so on. They should be taken into account properly to provide quality recommendation for users. In this work, we contribute a new solution named short for Neural Multi-Task Recommendation (NMTR) for learning recommender systems from user multi-behavior data. We develop a neural network model to capture the complicated and multi-type interactions between users and items. In particular, our model accounts for the cascading relationship among different types of behaviors (e.g., a user must click on a product before purchasing it). To fully exploit the signal in the data of multiple types of behaviors, we perform a joint optimization based on the multi-task learning framework, where the optimization on a behavior is treated as a task. Extensive experiments on two real-world datasets demonstrate that NMTR significantly outperforms state-of-the-art recommender systems that are designed to learn from both single-behavior data and multi-behavior data. Further analysis shows that modeling multiple behaviors is particularly useful for providing recommendation for sparse users that have very few interactions.

In recent years, much research effort on recommendation has been devoted to mining user behaviors, i.e., collaborative filtering, along with the general information which describes users or items, e.g., textual attributes, categorical demographics, product images, and so on. Price, an important factor in marketing - which determines whether a user will make the final purchase decision on an item - surprisingly, has received relatively little scrutiny. In this work, we aim at developing an effective method to predict user purchase intention with the focus on the price factor in recommender systems. The main difficulties are twofold: 1) the preference and sensitivity of a user on item price are unknown, which are only implicitly reflected in the items that the user has purchased, and 2) how the item price affects a user's intention depends largely on the product category, that is, the perception and affordability of a user on item price could vary significantly across categories. Towards the first difficulty, we propose to model the transitive relationship between user-to-item and item-to-price, taking the inspiration from the recently developed Graph Convolution Networks (GCN). The key idea is to propagate the influence of price on users with items as the bridge, so as to make the learned user representations be price-aware. For the second difficulty, we further integrate item categories into the propagation progress and model the possible pairwise interactions for predicting user-item interactions. We conduct extensive experiments on two real-world datasets, demonstrating the effectiveness of our GCN-based method in learning the price-aware preference of users. Further analysis reveals that modeling the price awareness is particularly useful for predicting user preference on items of unexplored categories.

The tutorial focuses on two major themes of recent advances in recommender systems: Part A: Recommendations in a Marketplace: Multi-sided marketplaces are steadily emerging as valuable ecosystems in many applications (e.g. Amazon, AirBnb, Uber), wherein the platforms have customers not only on the demand side (e.g. users), but also on the supply side (e.g. retailer). This tutorial focuses on designing search & recommendation frameworks that power such multi-stakeholder platforms. We discuss multi-objective ranking/recommendation techniques, discuss different ways in which stakeholders specify their objectives, highlight user specific characteristics (e.g. user receptivity) which could be leveraged when developing joint optimization modules and finally present a number of real world case-studies of such multi-stakeholder platforms.

Bundle recommendation aims to recommend a bundle of items for a user to consume as a whole. Existing solutions integrate user-item interaction modeling into bundle recommendation by sharing model parameters or learning in a multi-task manner, which cannot explicitly model the affiliation between items and bundles, and fail to explore the decision-making when a user chooses bundles. In this work, we propose a graph neural network model named BGCN (short forBundle Graph Convolutional Network ) for bundle recommendation. BGCN unifies user-item interaction, user-bundle interaction and bundle-item affiliation into a heterogeneous graph. With item nodes as the bridge, graph convolutional propagation between user and bundle nodes makes the learned representations capture the item level semantics. Through training based on hard-negative sampler, the user's fine-grained preferences for similar bundles are further distinguished. Empirical results on two real-world datasets demonstrate the strong performance gains of BGCN, which outperforms the state-of-the-art baselines by 10.77% to 23.18%.

Most existing recommender systems leverage users' complete original behavioral logs, which are collected from mobile devices and stored by the service provider and further fed into recommendation models. This may lead to a high risk of privacy leakage since the recommendation service provider may be trustless. Despite many research efforts on privacy-aware recommendation, the problem of building an effective recommender system completely preserving user privacy is still open.

Traditional recommendation models that usually utilize only one type of user-item interaction are faced with serious data sparsity or cold start issues. Multi-behavior recommendation taking use of multiple types of user-item interactions, such as clicks and favorites, can serve as an effective solution. Early efforts towards multi-behavior recommendation fail to capture behaviors' different influence strength on target behavior. They also ignore behaviors' semantics which is implied in multi-behavior data. Both of these two limitations make the data not fully exploited for improving the recommendation performance on the target behavior.

Most existing recommender systems leverage user behavior data of one type, such as the purchase behavior data in E-commerce. We argue that other types of user behavior data also provide valuable signal, such as views, clicks, and so on. In this work, we contribute a new solution named NMTR (short for Neural Multi-Task Recommendation) for learning recommender systems from user multi-behavior data. In particular, our model accounts for the cascading relationship among different types of behaviors (e.g., a user must click on a product before purchasing it). We perform a joint optimization based on the multi-task learning framework, where the optimization on a behavior is treated as a task. Extensive experiments on the real-world dataset demonstrate that NMTR significantly outperforms state-of-the-art recommender systems that are designed to learn from both single-behavior data and multi-behavior data.

Recommendation models mainly deal with categorical variables, such as user/item ID and attributes. Besides the high-cardinality issue, the interactions among such categorical variables are usually long-tailed, with the head made up of highly frequent values and a long tail of rare ones. This phenomenon results in the data sparsity issue, making it essential to regularize the models to ensure generalization. The common practice is to employ grid search to manually tune regularization hyperparameters based on the validation data. However, it requires non-trivial efforts and large computation resources to search the whole candidate space; even so, it may not lead to the optimal choice, for which different parameters should have different regularization strengths. In this paper, we propose a hyperparameter optimization method, lambdaOpt, which automatically and adaptively enforces regularization during training. Specifically, it updates the regularization coefficients based on the performance of validation data. With lambdaOpt, the notorious tuning of regularization hyperparameters can be avoided; more importantly, it allows fine-grained regularization (i.e. each parameter can have an individualized regularization coefficient), leading to better generalized models. We show how to employ lambdaOpt on matrix factorization, a classical model that is representative of a large family of recommender models. Extensive experiments on two public benchmarks demonstrate the superiority of our method in boosting the performance of top-K recommendation.

Social e-commerce, as a new concept of e-commerce, uses social media as a new prevalent platform for online shopping. Users are now able to view, add to cart, and buy products within a single social media app. In this paper, we address the problem of cross-platform recommendation for social e-commerce, i.e., recommending products to users when they are shopping through social media. To the best of our knowledge, this is a new and important problem for all e-commerce companies (e.g. Amazon, Alibaba), but has never been studied before.

Web systems that provide the same functionality usually share a certain amount of items. This makes it possible to combine data from different websites to improve recommendation quality, known as the cross-domain recommendation task. Despite many research efforts on this task, the main drawback is that they largely assume the data of different systems can be fully shared. Such an assumption is unrealistic - different systems are typically operated by different companies, and it may violate business privacy policy to directly share user behavior data since it is highly sensitive.

Social recommendation, which utilizes social relations to enhance recommender systems, has been gaining increasing attention recently with the rapid development of online social network. Existing social recommendation methods are based on the fact that users preference or decision is influenced by their social friends' behaviors. However, they assume that the influences of social relation are always the same, which violates the fact that users are likely to share preference on diverse products with different friends. In this paper, we present a novel CSR (short for C haracterized S ocial R egularization) model by designing a universal regularization term for modeling variable social influence. Our proposed model can be applied to both explicit and implicit iteration. Extensive experiments on a real-world dataset demonstrate that CSR significantly outperforms state-of-the-art social recommendation methods.

It is well-known that online services resort to various cookies to track users through users' online service identifiers (IDs) - in other words, when users access online services, various "fingerprints" are left behind in the cyberspace. As they roam around in the physical world while accessing online services via mobile devices, users also leave a series of "footprints" -- i.e., hints about their physical locations - in the physical world. This poses a potent new threat to user privacy: one can potentially correlate the "fingerprints" left by the users in the cyberspace with "footprints" left in the physical world to infer and reveal leakage of user physical world privacy, such as frequent user locations or mobility trajectories in the physical world - we refer to this problem as user physical world privacy leakage via user cyberspace privacy leakage. In this paper we address the following fundamental question: what kind - and how much - of user physical world privacy might be leaked if we could get hold of such diverse network datasets even without any physical location information. In order to conduct an in-depth investigation of these questions, we utilize the network data collected via a DPI system at the routers within one of the largest Internet operator in Shanghai, China over a duration of one month. We decompose the fundamental question into the three problems: i) linkage of various online user IDs belonging to the same person via mobility pattern mining; ii) physical location classification via aggregate user mobility patterns over time; and iii) tracking user physical mobility. By developing novel and effective methods for solving each of these problems, we demonstrate that the question of user physical world privacy leakage via user cyberspace privacy leakage is not hypothetical, but indeed poses a real potent threat to user privacy.