The attribution technique enhances the credibility of LLMs by adding citations to the generated sentences, enabling users to trace back to the original sources and verify the reliability of the output. However, existing instruction-tuned attributed LLMs often fail to properly interpret the contextual semantics of citation symbols (e.g., [i]) during text generation. This shortcoming arises from their insufficient awareness of the context information surrounding citation markers, which in turn leads to disjointed references and poor integration of retrieved knowledge into the generated content. To address this issue, we propose a novel Contextual-aware Citation generation framework (C²-Cite) that explicitly integrates the semantic relationships between citation markers and their referenced content. Specifically, a contextual citation alignment mechanism is adopted: it first encodes the retrieved document contexts into the symbol representation of citations, then aligns the marker numbers by decoding information from a citation router function. This mechanism enables the transformation of citation markers from generic placeholders into active knowledge pointers that link to the referenced source information. Experimental results on the ALCE benchmark across three datasets validate our framework C²-Cite++: it outperforms the SOTA baseline by an average of 5.8% in citation quality and 17.4% in response correctness. The implementation is publicly available at https://github.com/BAI-LAB/c2cite

In real-world industrial scenarios, large language models (LLMs) require Continuous Learning (CL) to adapt to diverse tasks as opera- tional requirements diversify, demanding self-evolution capabilities to autonomously refine their knowledge and adapt to dynamic envi- ronments. However, existing CL approaches, such as replay-based and parameter isolation techniques, struggle with the catastrophic forgetting problem: new task training degrades performance on prior tasks due to the model's adaptation to new data distributions, which weakens its generalization to old tasks. To address this issue, we propose a novel parameter-efficient adversarial MoE framework, MoE-CL, for industrial-scale self-evolving continual instruction tuning of LLMs. Specifically, MoE-CL employs a dual-expert archi- tecture to enable self-evolution: a dedicated LoRA expert for each task to preserve task-specific knowledge, ensuring parameter inde- pendence and mitigating forgetting, and a shared LoRA expert to facilitate cross-task knowledge transfer. Specifically, a task-aware discriminator within a Generative Adversarial Network (GAN) is integrated into the shared expert to suppress task-irrelevant noise, ensuring only task-aligned knowledge is transferred during se- quential task training. Through adversarial training, the shared ex- pert learns generalized representations that mimic the task-aware discriminator, while dedicated experts retain task-specific details, balancing knowledge retention and cross-task generalization—key to the model's self-evolution by autonomously optimizing knowl- edge integration across tasks. Extensive experiments on a public MTL5 benchmark and an industrial Tencent3 benchmark validate MoE-CL's effectiveness in self-evolving continual learning. In real- world A/B testing on content compliance review in the Tencent Video Platform, MoE-CL reduced manual review costs by 15.3%,demonstrating its applicability for large-scale industrial deploy- ment where self-evolution is critical for adapting to evolving op- erational demands. Implementation code is publicly available at https://github.com/BAI-LAB/MoE-CL.

With the expansion of business scenarios, real recommender systems are facing challenges in dealing with the constantly emerging new tasks in multi-task learning frameworks. In this article, we attempt to improve the generalization ability of multi-task recommendations when dealing with new tasks. A novel two-stage prompt-tuning MTL framework (MPT-Rec) is proposed to address task irrelevance and training efficiency problems in multi-task recommender systems. Specifically, we disentangle the task-specific and task-sharing information in the multi-task pre-training stage and then use task-aware prompts to transfer knowledge from other tasks to the new task effectively. By freezing parameters in the pre-training tasks, MPT-Rec solves the negative impacts that may be brought by the new task and greatly reduces the training costs. Extensive experiments on three real-world datasets show the effectiveness of our proposed multi-task learning framework. MPT-Rec achieves the best performance compared to the SOTA multi-task learning method on three real-world datasets. Besides, it maintains comparable model performance but vastly improves the training efficiency (i.e., with up to 10% parameters in the full-training way) in the new task learning. Our code is publicly available at https://github.com/BAI-LAB/MPT-Rec .

GNN-based recommender systems have shown their vulnerability to shilling attacks in recent studies. By conducting shilling attacks on recommender systems, the attackers aim to have homogeneous impacts on all users. However, such indiscriminate attacks suffer from a waste of resources because even if the target item is promoted to users who are not interested, they are unlikely to click on them. In this paper, we conduct targeted shilling attacks in GNN-based recommender systems. By automatically constructing the features and edges of the fake users, our proposed framework AutoAttack achieves accurate attacks on a specific group of users while minimizing the impact on non-target users. Specifically, the features of fake users are generated based on a similarity function, which is optimized according to the features of target users. The structure of fake users is learned by conducting spectral clustering on the target users based on their graph Laplacian matrix, which contains the degree and adjacency information that provides guidance to the edge generation of fake users. We conduct extensive experiments on four real-world datasets in different GNN-based RS and evaluate the performance of our method on the shilling attack and recommendation tasks comprehensively, showing the effectiveness and flexibility of our framework.

Learning effective high-order feature interactions is very crucial in the CTR prediction task. However, it is very time-consuming to calculate high-order feature interactions with massive features in online e-commerce platforms. Most existing methods manually design a maximal order and further filter out the useless interactions from them. Although they reduce the high computational costs caused by the exponential growth of high-order feature combinations, they still suffer from the degradation of model capability due to the suboptimal learning of the restricted feature orders. The solution to maintain the model capability and meanwhile keep it efficient is a technical challenge, which has not been adequately addressed. To address this issue, we propose an adaptive feature interaction learning model, named as EulerNet, in which the feature interactions are learned in a complex vector space by conducting space mapping according to Euler's formula. EulerNet converts the exponential powers of feature interactions into simple linear combinations of the modulus and phase of the complex features, making it possible to adaptively learn the high-order feature interactions in an efficient way. Furthermore, EulerNet incorporates the implicit and explicit feature interactions into a unified architecture, which achieves the mutual enhancement and largely boosts the model capabilities. Such a network can be fully learned from data, with no need of pre-designed form or order for feature interactions. Extensive experiments conducted on three public datasets have demonstrated the effectiveness and efficiency of our approach. Our code is available at: https://github.com/RUCAIBox/EulerNet.

With the growth of high-dimensional sparse data in web-scale recommender systems, the computational cost to learn high-order feature interaction in CTR prediction task largely increases, which limits the use of high-order interaction models in real industrial applications. Some recent knowledge distillation based methods transfer knowledge from complex teacher models to shallow student models for accelerating the online model inference. However, they suffer from the degradation of model accuracy in knowledge distillation process. It is challenging to balance the efficiency and effectiveness of the shallow student models. To address this problem, we propose a Directed Acyclic Graph Factorization Machine (KD-DAGFM) to learn the high-order feature interactions from existing complex interaction models for CTR prediction via Knowledge Distillation. The proposed lightweight student model DAGFM can learn arbitrary explicit feature interactions from teacher networks, which achieves approximately lossless performance and is proved by a dynamic programming algorithm. Besides, an improved general model KD-DAGFM+ is shown to be effective in distilling both explicit and implicit feature interactions from any complex teacher model. Extensive experiments are conducted on four real-world datasets, including a large-scale industrial dataset from WeChat platform with billions of feature dimensions. KD-DAGFM achieves the best performance with less than 21.5% FLOPs of the state-of-the-art method on both online and offline experiments, showing the superiority of DAGFM to deal with the industrial scale data in CTR prediction task.

Multi-Task Learning (MTL) has attracted increasing attention in recommender systems. A crucial challenge in MTL is to learn suitable shared parameters among tasks and to avoid negative transfer of information. The most recent sparse sharing models use independent parameter masks, which only activate useful parameters for a task, to choose the useful subnet for each task. However, as all the subnets are optimized in parallel for each task independently, it is faced with the problem of conflict between parameter gradient updates (i.e, parameter conflict problem). To address this challenge, we propose a novel Contrastive Sharing Recommendation model in MTL learning (CSRec). Each task in CSRec learns from the subnet by the independent parameter mask as in sparse sharing models, but a contrastive mask is carefully designed to evaluate the contribution of the parameter to a specific task. The conflict parameter will be optimized relying more on the task which is more impacted by the parameter. Besides, we adopt an alternating training strategy in CSRec, making it possible to self-adaptively update the conflict parameters by fair competitions. We conduct extensive experiments on three real-world large scale datasets, i.e., Tencent Kandian, Ali-CCP and Census-income, showing better effectiveness of our model over state-of-the-art methods for both offline and online MTL recommendation scenarios.

Video Question Answering (VideoQA) is a challenging video understanding task since it requires a deep understanding of both question and video. Previous studies mainly focus on extracting sophisticated visual and language embeddings, fusing them by delicate hand-crafted networks. However, the relevance of different frames, objects, and modalities to the question are varied along with the time, which is ignored in most of existing methods. Lacking understanding of the the dynamic relationships and interactions among objects brings a great challenge to VideoQA task. To address this problem, we propose a novel Relation-aware Hierarchical Attention (RHA) framework to learn both the static and dynamic relations of the objects in videos. In particular, videos and questions are embedded by pre-trained models firstly to obtain the visual and textual features. Then a graph-based relation encoder is utilized to extract the static relationship between visual objects. To capture the dynamic changes of multimodal objects in different video frames, we consider the temporal, spatial, and semantic relations, and fuse the multimodal features by hierarchical attention mechanism to predict the answer. We conduct extensive experiments on a large scale VideoQA dataset, and the experimental results demonstrate that our RHA outperforms the state-of-the-art methods.

In social recommendation, the purchase decision of users is influenced by their basic preference of items, as well as the social influence of peers. Such social connections had been proved to be effective in modeling users' preference of items. However, most models in social recommender literature only considered two types of relations, i.e., user-item relation in interaction network and user-user relation in social network. The temporal sequential information of items, i.e., item-item relation, can also be utilized to infer the preference of users, but had been ignored in almost all of the graph based recommendation models. Two issues of such temporal information had not been well studied in social recommender systems: the temporal strength information, i.e., the real purchase time of an item, and its influence on social relations. To address the above issues, we propose a novel Temporal Enhanced Graph Model for Social Recommendation (TGRec). In TGRec, the purchase time information between items is characterized as a special temporal relation, and the purchase decision of users depends on three factors: (1) a user's basic preference of items, (2) the collaborative influence of peers, (3) the temporal impact of previous items bought by the user. Experimental results on three real-world commerce datasets demonstrate the effectiveness of our model for social recommendation, showing the usefulness of modeling the temporal information in heterogeneous graph.

Applying reinforcement learning (RL) in recommender systems is attractive but costly due to the constraint of the interaction with real customers, where performing online policy learning through interacting with real customers usually harms customer experiences. A practical alternative is to build a recommender agent offline from logged data, whereas directly using logged data offline leads to the problem of selection bias between logging policy and the recommendation policy. The existing direct offline learning algorithms, such as Monte Carlo methods and temporal difference methods are either computationally expensive or unstable on convergence. To address these issues, we propose Pseudo Dyna-Q (PDQ). In PDQ, instead of interacting with real customers, we resort to a customer simulator, referred to as the World Model, which is designed to simulate the environment and handle the selection bias of logged data. During policy improvement, the World Model is constantly updated and optimized adaptively, according to the current recommendation policy. This way, the proposed PDQ not only avoids the instability of convergence and high computation cost of existing approaches but also provides unlimited interactions without involving real customers. Moreover, a proved upper bound of empirical error of reward function guarantees that the learned offline policy has lower bias and variance. Extensive experiments demonstrated the advantages of PDQ on two real-world datasets against state-of-the-arts methods.

Heterogeneous information network (HIN) contains multiple types of entities and relations. Most of existing HIN embedding methods learn the semantic information based on the heterogeneous structures between different entities, which are implicitly assumed to be complete. However, in real world, it is common that some relations are partially observed due to privacy or other reasons, resulting in a sparse network, in which the structure may be incomplete, and the ”unseen” links may also be positive due to the missing relations in data collection. To address this problem, we propose a novel and principled approach: a Multi-View Adversarial Completion Model (MV-ACM). Each relation space is characterized in a single viewpoint, enabling us to use the topological structural information in each view. Based on the multi-view architecture, an adversarial learning process is utilized to learn the reciprocity (i.e., complementary information) between different relations: In the generator, MV-ACM generates the complementary views by computing the similarity of the semantic representation of the same node in different views; while in the discriminator, MV-ACM discriminates whether the view is complementary by the topological structural similarity. Then we update the node’s semantic representation by aggregating neighborhoods information from the syncretic views. We conduct systematical experiments1 on six real-world networks from varied domains: AMiner, PPI, YouTube, Twitter, Amazon and Alibaba. Empirical results show that MV-ACM significantly outperforms the state-of-the-art approaches for both link prediction and node classification tasks.

In e-commerce, users' demands are not only conditioned by their profile and preferences, but also by their recent purchases that may generate new demands, as well as periodical demands that depend on purchases made some time ago. We call them respectively short-term demands and long-term demands. In this paper, we propose a novel self-attentive Continuous-Time Recommendation model (CTRec) for capturing the evolving demands of users over time. For modeling such time-sensitive demands, a Demand-aware Hawkes Process (DHP) framework is designed in CTRec to learn from the discrete purchase records of users. More specifically, a convolutional neural network is utilized to capture the short-term demands; and a self-attention mechanism is employed to capture the periodical purchase cycles of long-term demands. All types of demands are fused in DHP to make final continuous-time recommendations. We conduct extensive experiments on four real-world commercial datasets to demonstrate that CTRec is effective for general sequential recommendation problems, including next-item and next-session/basket recommendations. We observe in particular that CTRec is capable of learning the purchase cycles of products and estimating the purchase time of a product given a user.

Presents corrections to author information from the paper, “Characterizing and predicting early reviewers for effective product marketing on e-commerce websites,” (Bai, T., et al), IEEE Trans. Knowl. Data Eng., vol. 30, no. 12, pp. 2271–2284, Dec. 2018.

Next basket recommendation is a new type of recommendation, which recommends a set of items, or a basket, to the user. Purchase in basket is a common behavior of consumers. Recently, deep neural networks have been applied to model sequential transactions of baskets in next basket recommendation. However, current methods do not track the user's evolving appetite for items explicitly, and they ignore important item attributes such as product category. In this paper, we propose a novel Attribute-aware Neural Attentive Model (ANAM) to address these problems. ANAM adopts an attention mechanism to explicitly model user's evolving appetite for items, and utilizes a hierarchical architecture to incorporate the attribute information. In specific, ANAM utilizes a recurrent neural network to model the user's sequential behavior over time, and relays the user's appetite toward items and their attributes to next basket through attention weights shared across baskets on the two different hierarchies. Experiment results on two public datasets (ıe Ta-Feng and JingDong) demonstrate the effectiveness of our ANAM model for next basket recommendation.

Online reviews have become an important source of information for users before making an informed purchase decision. Early reviews of a product tend to have a high impact on the subsequent product sales. In this paper, we take the initiative to study the behavior characteristics of early reviewers through their posted reviews on two real-world large e-commerce platforms, i.e., Amazon and Yelp. In specific, we divide product lifetime into three consecutive stages, namelyearly,majority, andlaggards. A user who has posted a review in the early stage is considered as an early reviewer. We quantitatively characterize early reviewers based on their rating behaviors, the helpfulness scores received from others and the correlation of their reviews with product popularity. We have found that (1) an early reviewer tends to assign a higher average rating score; and (2) an early reviewer tends to post more helpful reviews. Our analysis of product reviews also indicates that early reviewers’ ratings and their received helpfulness scores are likely to influence product popularity. By viewing the review posting process as a multiplayer competition game, we propose a novel margin-based embedding model for early reviewer prediction. Extensive experiments on two different e-commerce datasets have shown that our proposed approach outperforms a number of competitive baselines.

Recently, deep neural networks have been widely applied to recommender systems. A representative work is to utilize deep learning for modeling complex user-item interactions. However, similar to traditional latent factor models by factorizing user-item interactions, they tend to be ineffective to capture localized information. Localized information, such as neighborhood, is important to recommender systems in complementing the user-item interaction data. Based on this consideration, we propose a novel Neighborhood-based Neural Collaborative Filtering model (NNCF). To the best of our knowledge, it is the first time that the neighborhood information is integrated into the neural collaborative filtering methods. Extensive experiments on three real-world datasets demonstrate the effectiveness of our model for the implicit recommendation task.