Large language models have given rise to interactive recommendation agents (IRAs). Through proactive clarification, tool invocation, and dynamic dialogue, IRAs shift recommender systems from passive prediction to interactive, proactive intelligence. For training IRAs, agentic reinforcement learning offers a natural pathway, as it enables models to learn interactive capabilities directly from environmental feedback without requiring costly annotated data. However, this process faces three key challenges: credit assignment in multi-turn interactions, efficient exploration in large action spaces, and coordinated learning of multiple interactive skills.

Jinfeng Zhou, Yuxuan Chen, Yihan Shi, Xuanming Zhang, Leqi Lei, Yi Feng, Zexuan Xiong, Miao Yan, Xunzhi Wang, Yaru Cao, Jianing Yin, Shuai Wang, Quanyu Dai, Zhenhua Dong, Hongning Wang, Minlie Huang. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Large Language Models (LLMs) have demonstrated remarkable capabilities in language understanding and commonsense reasoning, yet they often struggle with constraint satisfaction in planning problems. Previous studies relying on test-time improvement with self-evaluation fail to address this limitation effectively. In this work, we identify this critical gap and propose a novel neuro-symbolic framework, Reinforced Neuro-Symbolic Planning (\algo), that enhances LLM-powered planning by incorporating a symbolic verifier. The verifier provides explicit feedback on constraint satisfaction, enabling iterative refinement of the state evaluation. Specifically, we utilize the outcome feedback from each logical goal to update the process value along planning paths through a reinforcement value function maximization objective. We further employ T-norms to aggregate the satisfaction levels of multiple constraints, which provided more effective guidance for the test-time search. Our framework bridges the strengths of neural and symbolic methods, leveraging the generative power of LLMs while ensuring rigorous adherence to constraints through symbolic verification. Extensive experiments demonstrate that our approach significantly improves planning accuracy and constraint satisfaction across various domains, outperforming traditional self-evaluation methods. It highlights the potential of hybrid neuro-symbolic systems to address complex constrained planning tasks.

Recent advances in retrieval-augmented generation (RAG) furnish large language models (LLMs) with iterative retrievals of relevant information to handle complex multi-hop questions. These methods typically alternate between LLM reasoning and retrieval to accumulate external information into the LLM's context. However, the ever-growing context inherently imposes an increasing burden on the LLM to perceive connections among critical information pieces, with futile reasoning steps further exacerbating this overload issue. In this paper, we present KnowTrace, an elegant RAG framework to (1) mitigate the context overload and (2) bootstrap higher-quality multi-step reasoning. Instead of simply piling the retrieved contents, KnowTrace autonomously traces out desired knowledge triplets to organize a specific knowledge graph relevant to the input question. Such a structured workflow not only empowers the LLM with an intelligible context for inference, but also naturally inspires a reflective mechanism of knowledge backtracing to identify contributive LLM generations as process supervision data for self-bootstrapping. Extensive experiments show that KnowTrace consistently surpasses existing methods across three multi-hop question answering benchmarks, and the bootstrapped version further amplifies the gains.

Xueyang Feng, Bo Lan, Quanyu Dai, Lei Wang, Jiakai Tang, Xu Chen, Zhenhua Dong, Ji-Rong Wen. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

LLMs have emerged as powerful evaluators in the LLM-as-a-Judge paradigm, offering significant efficiency and flexibility compared to human judgments. However, previous methods primarily rely on single-point evaluations, overlooking the inherent diversity and uncertainty in human evaluations. This approach leads to information loss and decreases the reliability of evaluations. To address this limitation, we propose a novel training framework that explicitly aligns the LLM-generated judgment distribution with human evaluation distributions. Specifically, we propose a distributional alignment objective based on KL divergence, combined with an auxiliary cross-entropy regularization to stabilize the training process. Furthermore, due to limited human annotations, empirical human distributions are merely noisy estimates of the true underlying distribution. We therefore incorporate adversarial training to ensure a robust alignment with this true distribution, rather than overfitting to its imperfect approximation. Extensive experiments across various LLM backbones and evaluation tasks demonstrate that our framework significantly outperforms existing closed-source LLMs and conventional single-point alignment methods, with superior alignment quality, strong robustness, and competitive evaluation accuracy.

Current Large Language Models (LLMs) are confronted with overwhelming information volume when comprehending long-form documents. This challenge raises the imperative of a cohesive memory module, which can elevate vanilla LLMs into autonomous reading agents. Despite the emergence of some heuristic approaches, a systematic design principle remains absent. To fill this void, we draw inspiration from Jean Piaget's Constructivist Theory, illuminating three traits of the agentic memory---structured schemata, flexible assimilation, and dynamic accommodation. This blueprint forges a clear path toward a more robust and efficient memory system for LLM-based reading comprehension. To this end, we develop CAM, a prototype implementation of Constructivist Agentic Memory that simultaneously embodies the structurality, flexibility, and dynamicity. At its core, CAM is endowed with an incremental overlapping clustering algorithm for structured memory development, supporting both coherent hierarchical summarization and online batch integration. During inference, CAM adaptively explores the memory structure to activate query-relevant information for contextual response, akin to the human associative process. Compared to existing approaches, our design demonstrates dual advantages in both performance and efficiency across diverse long-text reading comprehension tasks, including question answering, query-based summarization, and claim verification.

LLM-based agents have been widely applied as personal assistants, capable of memorizing information from user messages and responding to personal queries. However, there still lacks an objective and automatic evaluation on their memory capability, largely due to the challenges in constructing reliable questions and answers (QAs) according to user messages. In this paper, we propose MemSim, a Bayesian simulator designed to automatically construct reliable QAs from generated user messages, simultaneously keeping their diversity and scalability. Specifically, we introduce the Bayesian Relation Network (BRNet) and a causal generation mechanism to mitigate the impact of LLM hallucinations on factual information, facilitating the automatic creation of an evaluation dataset. Based on MemSim, we generate a dataset in the daily-life scenario, named MemDaily, and conduct extensive experiments to assess the effectiveness of our approach. We also provide a benchmark for evaluating different memory mechanisms in LLM-based agents with the MemDaily dataset.

In online advertising, uncertainty calibration aims to adjust a ranking model's probability predictions to better approximate the true likelihood of an event, e.g., a click or a conversion. However, existing calibration approaches may lack the ability to effectively model complex nonlinear relations, consider context features, and achieve balanced performance across different data subsets. To tackle these challenges, we introduce a novel model called Monotonic Calibration Networks, featuring three key designs: a monotonic calibration function (MCF), an order-preserving regularizer, and a field-balance regularizer. The nonlinear MCF is capable of naturally modeling and universally approximating the intricate relations between uncalibrated predictions and the posterior probabilities, thus being much more expressive than existing methods. MCF can also integrate context features using a flexible model architecture, thereby achieving context awareness. The order-preserving and field-balance regularizers promote the monotonic relationship between adjacent bins and the balanced calibration performance on data subsets, respectively. Experimental results on both public and industrial datasets demonstrate the superior performance of our method in generating well-calibrated probability predictions.

Large language models (LLMs) are increasingly leveraged as foundational backbones in the development of advanced recommender systems, offering enhanced capabilities through their extensive knowledge and reasoning.Existing llm-based recommender systems (RSs) often face challenges due to the significant differences between the linguistic semantics of pre-trained LLMs and the collaborative semantics essential for RSs.These systems use pre-trained linguistic semantics but learn collaborative semantics from scratch via the llm-Backbone.However, LLMs are not designed for recommendations, leading to inefficient collaborative learning, weak result correlations, and poor integration of traditional RS features.To address these challenges, we propose EAGER-LLM, a decoder-only llm-based generative recommendation framework that integrates endogenous and exogenous behavioral and semantic information in a non-intrusive manner.Specifically, we propose 1) dual-source knowledge-rich item indices that integrates indexing sequences * Both authors contributed equally to this research.

Large language model (LLM)-based agents have recently attracted much attention from the research and industry communities. Compared with original LLMs, LLM-based agents are featured in their self-evolving capability, which is the basis for solving real-world problems that need long-term and complex agent-environment interactions. The key component to support agent-environment interactions is the memory of the agents. While previous studies have proposed many promising memory mechanisms, they are scattered in different papers, and there lacks a systematical review to summarize and compare these works from a holistic perspective, failing to abstract common and effective designing patterns for inspiring future studies. To bridge this gap, in this article, we propose a comprehensive survey on the memory mechanism of LLM-based agents. In specific, we first discuss “what is” and “why do we need” the memory in LLM-based agents. Then, we systematically review previous studies on how to design and evaluate the memory module. In addition, we also present many agent applications, where the memory module plays an important role. At last, we analyze the limitations of existing work and show important future directions. To keep up with the latest advances in this field, we create a repository at https://github.com/nuster1128/LLM_Agent_Memory_Survey .

Recommender systems have a significant impact on various real-world applications, shaping people's daily lives and enhancing productivity. Traditional recommender models aim to collect extensive user information to accurately estimate user preferences. However, in practical scenarios, users may not want all their behaviors to be included in the model training process. This paper introduces a novel recommendation paradigm that allows users to indicate their ``willingness'' regarding which data should contribute to model training. The models are then optimized to maximize utility, which considers the trade-off between recommendation performance and respecting user preferences. The recommendation problem is formulated as a multiplayer game, with each user acting as a player and using a selection vector to indicate their willingness to include specific interacted items in training. To efficiently solve this game, an influence function-based model is proposed to approximate recommendation performances for different actions without re-optimizing the model. Furthermore, an enhanced model leveraging multiple anchor actions for the influence function is introduced to improve performance approximation accuracy. The convergence rate of the algorithm is theoretically analyzed, and the advantages of incorporating multiple anchor actions are demonstrated. Extensive experiments on both simulated and real-world datasets validate the effectiveness of the proposed models in balancing recommendation quality and user willingness. To promote this research direction, we have released our project at https://paitesanshi.github.io/IFRQE/.

Learning high-quality item embeddings is crucial for recommendation tasks such as matching and ranking. However, existing methods often rely on ID-based item embeddings learned end-to-end with downstream recommendation models, which may suffer from overfitting and limited generalizability. In this paper, we aim to learn universal item embeddings (dubbed UniEmbedding) that capture multi-modal semantics, generalize across multiple domains, and serve different downstream tasks. To achieve this goal, we introduce the UniEmbedding pretraining framework, which includes three modules: a domain-aware multi-modal adapter, a user-view projection module, and contrastive learning objectives across domains. Compared to naive ID embeddings, UniEmbedding provides rich semantic information that generalizes more effectively across domains. Unlike multi-modal embeddings directly extracted from off-the-shelf pretrained models, UniEmbedding achieves better alignment between content semantics and behaviors. We evaluated UniEmbedding on both public and industrial datasets, demonstrating its effectiveness in matching and ranking tasks. Furthermore, UniEmbedding has been deployed in multiple recommendation applications at Huawei, resulting in significant gains in user engagement metrics.

Explainable recommendation has gained significant attention due to its potential to enhance user trust and system transparency. Previous studies primarily focus on refining model architectures to generate more informative explanations, assuming that the explanation data is sufficient and easy to acquire. However, in practice, obtaining the ground truth for explanations can be costly since individuals may not be inclined to put additional efforts to provide behavior explanations. In this paper, we study a novel problem in the field of explainable recommendation, that is, “given a limited budget to incentivize users to provide behavior explanations, how to effectively collect data such that the downstream models can be better optimized?” To solve this problem, we propose an active learning framework for recommender system, which consists of an acquisition function for sample collection and an explainable recommendation model to provide the final results. We consider both uncertainty and influence based strategies to design the acquisition function, which can determine the sample effectiveness from complementary perspectives. To demonstrate the effectiveness of our framework, we conduct extensive experiments based on real-world datasets.

With the popularity of online music services, personalized music recommendation has garnered much research interest. Recommendation models are typically trained on datasets constructed from user feedback, which includes both the active feedback (e.g., clicking the Like or Skip buttons) and passive feedback (e.g., auto-play), with passive feedback comprising the majority. Due to the unavailability of user attention, the massive amount of passive feedback is unreliable, significantly compromising the quality of the training data. How to estimate the user's attention on the target music has become a critical problem in music recommendation. Heuristic methods such as exponential decay and negative sampling have been proposed. However, they either neglect the sequential dependencies between feedback actions or utilize only a small fraction of passive samples, leading to inaccurate and biased attention estimation. In this paper, we naturally propose modeling user attention prediction as a positive-unlabeled (PU) learning problem, where active feedback is treated as positive samples and passive feedback is treated as unlabeled samples, as we can only ensure that the user's attention is focused when she provides active feedback. Then we propose an extended PU-learning model with sequential dependencies, called UAE, which contains an unbiased user attention estimator and an unbiased propensity estimator. Subsequently, a joint learning algorithm is developed in which the attention and propensity estimators are optimized in alternating fashion. Theoretical analysis shows the unbiasedness and variance of the attention estimator and the propensity estimator. Extensive experiments on two large-scale datasets demonstrate the proposed UAE's effectiveness and generality in enhancing downstream music recommendation. One week online A/B testing on Huawei Music App manifests that UAE can significantly increase the users' play count and time over 2%, further demonstrating the effectiveness of UAE in real-world music recommendation products.

Personalized recommendation serves as a ubiquitous channel for users to discover information tailored to their interests. However, traditional recommendation models primarily rely on unique IDs and categorical features for user-item matching, potentially overlooking the nuanced essence of raw item contents across multiple modalities such as text, image, audio, and video. This underutilization of multimodal data poses a limitation to recommender systems, especially in multimedia services like news, music, and short-video platforms. The recent advancements in large multimodal models offer new opportunities and challenges in developing content-aware recommender systems. This survey seeks to provide a comprehensive exploration of the latest advancements and future trajectories in multimodal pretraining, adaptation, and generation techniques, as well as their applications in enhancing recommender systems. Furthermore, we discuss current open challenges and opportunities for future research in this dynamic domain. We believe that this survey, alongside the curated resources, will provide valuable insights to inspire further advancements in this evolving landscape.

Music source separation and pitch estimation are two vital tasks in music information retrieval. Typically, the input of pitch estimation is obtained from the output of music source separation. Therefore, existing methods have tried to perform these two tasks simultaneously, so as to leverage the mutually beneficial relationship between both tasks. However, these methods still face two critical challenges that limit the improvement of both tasks: the lack of labeled data and joint learning optimization. To address these challenges, we propose a Model-Agnostic Joint Learning (MAJL) framework for both tasks. MAJL is a generic framework and can use variant models for each task. It includes a two-stage training method and a dynamic weighting method named Dynamic Weights on Hard Samples (DWHS), which addresses the lack of labeled data and joint learning optimization, respectively. Experimental results on public music datasets show that MAJL outperforms state-of-the-art methods on both tasks, with significant improvements of 0.92 in Signal-to-Distortion Ratio (SDR) for music source separation and 2.71% in Raw Pitch Accuracy (RPA) for pitch estimation. Furthermore, comprehensive studies not only validate the effectiveness of each component of MAJL, but also indicate the great generality of MAJL in adapting to different model architectures.

Benefiting from the powerful language expression and planning capabilities of Large Language Models (LLMs), LLM-based autonomous agents have achieved promising performance in various downstream tasks. Recently, based on the development of single-agent systems, researchers propose to construct LLM-based multi-agent systems to tackle more complicated tasks. In this paper, we propose a novel framework, named COPPER, to enhance the collaborative capabilities of LLM-based agents with the self-reflection mechanism. To improve the quality of reflections, we propose to fine-tune a shared reflector, which automatically tunes the prompts of actor models using our counterfactual PPO mechanism. On the one hand, we propose counterfactual rewards to assess the contribution of a single agent’s reflection within the system, alleviating the credit assignment problem. On the other hand, we propose to train a shared reflector, which enables the reflector to generate personalized reflections according to agent roles, while reducing the computational resource requirements and improving training stability. We conduct experiments on three datasets to evaluate the performance of our model in multi-hop question answering, mathematics, and chess scenarios. Experimental results show that COPPER possesses stronger reflection capabilities and exhibits excellent generalization performance across different actor models.

In recommender systems, a common problem is the presence of various biases in the collected data, which deteriorates the generalization ability of the recommendation models and leads to inaccurate predictions. Doubly robust (DR) learning has been studied in many tasks in RS, with the advantage that unbiased learning can be achieved when either a single imputation or a single propensity model is accurate. In this paper, we propose a multiple robust (MR) estimator that can take the advantage of multiple candidate imputation and propensity models to achieve unbiasedness. Specifically, the MR estimator is unbiased when any of the imputation or propensity models, or a linear combination of these models is accurate. Theoretical analysis shows that the proposed MR is an enhanced version of DR when only having a single imputation and propensity model, and has a smaller bias. Inspired by the generalization error bound of MR, we further propose a novel multiple robust learning approach with stabilization. We conduct extensive experiments on real-world and semi-synthetic datasets, which demonstrates the superiority of the proposed approach over state-of-the-art methods.

The key of sequential recommendation lies in the accurate item correlation modeling. Previous models infer such information based on item co-occurrences, which may fail to capture the real causal relations, and impact the recommendation performance and explainability. In this paper, we equip sequential recommendation with a novel causal discovery module to capture causalities among user behaviors. Our general idea is firstly assuming a causal graph underlying item correlations, and then we learn the causal graph jointly with the sequential recommender model by fitting the real user behavior data. More specifically, in order to satisfy the causality requirement, the causal graph is regularized by a differentiable directed acyclic constraint. Considering that the number of items in recommender systems can be very large, we represent different items with a unified set of latent clusters, and the causal graph is defined on the cluster level, which enhances the model scalability and robustness. In addition, we provide theoretical analysis on the identifiability of the learned causal graph. To the best of our knowledge, this paper makes a first step towards combining sequential recommendation with causal discovery. For evaluating the recommendation performance, we implement our framework with different neural sequential architectures, and compare them with many state-of-the-art methods based on real-world datasets. Empirical studies manifest that our model can on average improve the performance by about 6.1% and 11.3% on F1and NDCG, respectively. To evaluate the model explainability, we build a new dataset with human labeled explanations for both quantitative and qualitative analysis.

Explainable recommendation has attracted much attention from the industry and academic communities. It has shown great potential to improve the recommendation persuasiveness, informativeness and user satisfaction. In the past few years, while a lot of promising explainable recommender models have been proposed, the datasets used to evaluate them still suffer from several limitations, for example, the explanation ground truths are not labeled by the real users, the explanations are mostly single-modal and around only one aspect. To bridge these gaps, in this paper, we build a new explainable recommendation dataset, which, to our knowledge, is the first contribution that provides a large amount of real user labeled multi-modal and multi-aspect explaination ground truths. In specific, we firstly develop a video recommendation platform, where a series of questions around the recommendation explainability are carefully designed. Then, we recruit about 3000 high-quality labelers with different backgrounds to use the system, and collect their behaviors and feedback to our questions. In this paper, we detail the construction process of our dataset and also provide extensive analysis on its characteristics. In addition, we develop a library, where ten well-known explainable recommender models are implemented in a unified framework. Based on this library, we build several benchmarks for different explainable recommendation tasks. At last, we present many new opportunities brought by our dataset, which are expected to promote the field of explainable recommendation. Our dataset, library and the related documents have been released at https://reasoner2023.github.io/.

Estimating individual treatment effects from observational data is challenging due to treatment selection bias. Prevalent methods mainly mitigate this issue by aligning different treatment groups in the latent space, the core of which is the calculation of distribution discrepancy. However, two issues that are often overlooked can render these methods invalid: (1) mini-batch sampling effects (MSE), where the calculated discrepancy is erroneous in non-ideal mini-batches with outcome imbalance and outliers; (2) unobserved confounder effects (UCE), where the unobserved confounders are not considered in the discrepancy calculation. Both of these issues invalidate the calculated discrepancy, mislead the training of estimators, and thus impede the handling of treatment selection bias. To tackle these issues, we propose Entire Space CounterFactual Regression (ESCFR), which is a new take on optimal transport technology in the context of causality. Specifically, based on the canonical optimal transport framework, we propose a relaxed mass-preserving regularizer to address the MSE issue and design a proximal factual outcome regularizer to handle the UCE issue. Extensive experiments demonstrate that ESCFR estimates distribution discrepancy accurately, handles the treatment selection bias effectively, and outperforms prevalent competitors significantly.

Recently, there has been growing interest in integrating causal inference into recommender systems to answer the hypothetical question: "what would be the potential feedback when a user is recommended a product?" Various unbiased estimators, including Inverse Propensity Score (IPS) and Doubly Robust (DR), have been proposed to address this question. However, these estimators often assume that confounders are precisely observable, which is not always the case in real-world scenarios. To address this challenge, we propose a novel method called Adversarial Training-based IPS (AT-IPS), which uses adversarial training to handle noisy confounders. The proposed method defines a feasible region for the confounders, obtains the worst-case noise (adversarial noise) within the region, and jointly trains the propensity model and the prediction model against such noise to improve their robustness. We provide a theoretical analysis of the accuracy-robustness tradeoff of AT-IPS and demonstrate its superior performance compared to other popular estimators on both real-world and semi-synthetic datasets.

Multi-layer perceptron (MLP) serves as a core component in many deep models for click-through rate (CTR) prediction. However, vanilla MLP networks are inefficient in learning multiplicative feature interactions, making feature interaction learning an essential topic for CTR prediction. Existing feature interaction networks are effective in complementing the learning of MLPs, but they often fall short of the performance of MLPs when applied alone. Thus, their integration with MLP networks is necessary to achieve improved performance. This situation motivates us to explore a better alternative to the MLP backbone that could potentially replace MLPs. Inspired by factorization machines, in this paper, we propose FINAL, a factorized interaction layer that extends the widely-used linear layer and is capable of learning 2nd-order feature interactions. Similar to MLPs, multiple FINAL layers can be stacked into a FINAL block, yielding feature interactions with an exponential degree growth. We unify feature interactions and MLPs into a single FINAL block and empirically show its effectiveness as a replacement for the MLP block. Furthermore, we explore the ensemble of two FINAL blocks as an enhanced two-stream CTR model, setting a new state-of-the-art on open benchmark datasets. FINAL can be easily adopted as a building block and has achieved business metric gains in multiple applications at Huawei. Our source code will be made available at MindSpore/models and FuxiCTR/model_zoo.

Explainable recommendation has recently attracted increasing attention from both academic and industry communities. Among different explainable strategies, generating natural language explanations is an important method, which can deliver more informative, flexible and readable explanations to facilitate better user decisions. Despite the effectiveness, existing models are mostly optimized based on the observed datasets, which can be skewed due to the selection or exposure bias. To alleviate this problem, in this paper, we formulate the task of explainable recommendation with a causal graph, and design a causality enhanced framework to generate unbiased explanations. More specifically, we firstly define an ideal unbiased learning objective, and then derive a tractable loss for the observational data based on the inverse propensity score (IPS), where the key is a sample re-weighting strategy for equalizing the loss and ideal objective in expectation. Considering that the IPS estimated from the sparse and noisy recommendation datasets can be inaccurate, we introduce a fault tolerant mechanism by minimizing the maximum loss induced by the sample weights near the IPS. For more comprehensive modeling, we further analyze and infer the potential latent confounders induced by the complex and diverse user personalities. We conduct extensive experiments by comparing with the state-of-the-art methods based on three real-world datasets to demonstrate the effectiveness of our method.

The high cost of data labeling often results in node label shortage in real applications. To improve node classification accuracy, graph-based semi-supervised learning leverages the ample unlabeled nodes to train together with the scarce available labeled nodes. However, most existing methods require the information of all nodes, including those to be predicted, during model training, which is not practical for dynamic graphs with newly added nodes. To address this issue, an adversarially regularized graph attention model is proposed to classify newly added nodes in a partially labeled graph. An attention-based aggregator is designed to generate the representation of a node by aggregating information from its neighboring nodes, thus naturally generalizing to previously unseen nodes. In addition, adversarial training is employed to improve the model’s robustness and generalization ability by enforcing node representations to match a prior distribution. Experiments on real-world datasets demonstrate the effectiveness of the proposed method in comparison with the state-of-the-art methods. The code is available at https://github.com/JiarenX/AGAIN.

This paper studies the problem of cross-network node classification to overcome the insufficiency of labeled data in a single network. It aims to leverage the label information in a partially labeled source network to assist node classification in a completely unlabeled or partially labeled target network. Existing methods for single network learning cannot solve this problem due to the domain shift across networks. Some multi-network learning methods heavily rely on the existence of cross-network connections, thus are inapplicable for this problem. To tackle this problem, we propose a novel graph transfer learning framework AdaGCN by leveraging the techniques of adversarial domain adaptation and graph convolution. It consists of two components: a semi-supervised learning component and an adversarial domain adaptation component. The former aims to learn class discriminative node representations with given label information of the source and target networks, while the latter contributes to mitigating the distribution divergence between the source and target domains to facilitate knowledge transfer. Extensive empirical evaluations on real-world datasets show that AdaGCN can successfully transfer class information with a low label rate on the source network and a substantial divergence between the source and target domains.

Music recommendation is usually modeled as a Click-Through Rate (CTR) prediction problem, which estimates the probability of a user listening a recommended song. CTR prediction can be formulated as a binary classification problem where the played songs are labeled as positive samples and the skipped songs are labeled as negative samples. However, such naively defined labels are noisy and biased in practice, causing inaccurate model predictions. In this work, we first identify serious label noise issues in an industrial music App, and then propose an adaptive Label Correction method for Denoising (LCD) music recommendation by ensembling the noisy labels and the model outputs to encourage a consensus prediction. Extensive offline experiments are conducted to evaluate the effectiveness of LCD on both industrial and public datasets. Furthermore, in a one-week online AB test, LCD also significantly increases both the music play count and time per user by 1% to 5%.

Understanding news content is critical to improving the quality of news recommendation. To achieve this goal, recent studies have attempted to apply pre-trained language models (PLMs) such as BERT for semantic-enhanced news recommendation. Despite their great success in offline evaluation, it is still a challenge to apply such large PLMs in real-time ranking model due to the stringent requirement in inference and updating time. To bridge this gap, we propose a plug-and-play pre-trainer, namely PREC, to learn both user and news encoders through multi-task pre-training. Instead of directly leveraging sophisticated PLMs for end-to-end inference, we focus on how to use the derived user and item representations to boost the performance of conventional lightweight models for click-through-rate prediction. This enables efficient online inference as well as compatibility to conventional models, which would significantly ease the practical deployment. We validate the effectiveness of PREC through both offline evaluation on public datasets and online A/B testing in an industrial application.

Recently, recommender system (RS) based on causal inference has gained much attention in the industrial community, as well as the states of the art performance in many prediction and debiasing tasks. Nevertheless, a unified causal analysis framework has not been established yet. Many causal-based prediction and debiasing studies rarely discuss the causal interpretation of various biases and the rationality of the corresponding causal assumptions. In this paper, we first provide a formal causal analysis framework to survey and unify the existing causal-inspired recommendation methods, which can accommodate different scenarios in RS. Then we propose a new taxonomy and give formal causal definitions of various biases in RS from the perspective of violating the assumptions adopted in causal analysis. Finally, we formalize many debiasing and prediction tasks in RS, and summarize the statistical and machine learning-based causal estimation methods, expecting to provide new research opportunities and perspectives to the causal RS community.

Post-click conversion rate (CVR) prediction is an essential task for discovering user interests and increasing platform revenues in a range of industrial applications. One of the most challenging problems of this task is the existence of severe selection bias caused by the inherent self-selection behavior of users and the item selection process of systems. Currently, doubly robust (DR) learning approaches achieve the state-of-the-art performance for debiasing CVR prediction. However, in this paper, by theoretically analyzing the bias, variance and generalization bounds of DR methods, we find that existing DR approaches may have poor generalization caused by inaccurate estimation of propensity scores and imputation errors, which often occur in practice. Motivated by such analysis, we propose a generalized learning framework that not only unifies existing DR methods, but also provides a valuable opportunity to develop a series of new debiasing techniques to accommodate different application scenarios. Based on the framework, we propose two new DR methods, namely DR-BIAS and DR-MSE. DR-BIAS directly controls the bias of DR loss, while DR-MSE balances the bias and variance flexibly, which achieves better generalization performance. In addition, we propose a novel tri-level joint learning optimization method for DR-MSE in CVR prediction, and an efficient training algorithm correspondingly. We conduct extensive experiments on both real-world and semi-synthetic datasets, which validate the effectiveness of our proposed methods.

In streaming media applications, like music Apps, songs are recommended in a continuous way in users' daily life. The recommended songs are played automatically although users may not pay any attention to them, posing a challenge of user attention bias in training recommendation models, i.e., the training instances contain a large number of false-positive labels (users' feedback). Existing approaches either directly use the auto-feedbacks or heuristically delete the potential false-positive labels. Both of the approaches lead to biased results because the false-positive labels cause the shift of training data distribution, hurting the accuracy of the recommendation models. In this paper, we propose a learning-based counterfactual approach to adjusting the user auto-feedbacks and learning the recommendation models using Neural Dueling Bandit algorithm, called NDB. Specifically, NDB maintains two neural networks: a user attention network for computing the importance weights that are used for modifying the original rewards, and another random network trained with dueling bandit for conducting online recommendations based on the modified rewards. Theoretical analysis showed that the modified rewards are statistically unbiased, and the learned bandit policy enjoys a sub-linear regret bound. Experimental results demonstrated that NDB can significantly outperform the state-of-the-art baselines.

Deep learning-based recommendation has become a widely adopted technique in various online applications. Typically, a deployed model undergoes frequent re-training to capture users' dynamic behaviors from newly collected interaction logs. However, the current model training process only acquires users' feedbacks as labels, but fails to take into account the errors made in previous recommendations. Inspired by the intuition that humans usually reflect and learn from mistakes, in this paper, we attempt to build a self-correction continual learning loop (dubbed ReLoop) for recommender systems. In particular, a new customized loss is employed to encourage every new model version to reduce prediction errors over the previous model version during training. Our ReLoop learning framework enables a continual self-correction process in the long run and thus is expected to obtain better performance over existing training strategies. Both offline experiments and an online A/B test have been conducted to validate the effectiveness of ReLoop.

The past two decades have witnessed the rapid development of personalized recommendation techniques. Despite the significant progress made in both research and practice of recommender systems, to date, there is a lack of a widely-recognized benchmarking standard in this field. Many of the existing studies perform model evaluations and comparisons in an ad-hoc manner, for example, by employing their own private data splits or using a different experimental setting. However, such conventions not only increase the difficulty in reproducing existing studies, but also lead to inconsistent experimental results among them. This largely limits the credibility and practical value of research results in this field. To tackle these issues, we present an initiative project aimed for open benchmarking for recommender systems. In contrast to some earlier attempts towards this goal, we take one further step by setting up a standardized benchmarking pipeline for reproducible research, which integrates all the details about datasets, source code, hyper-parameter settings, running logs, and evaluation results. The benchmark is designed with comprehensiveness and sustainability in mind. It spans both matching and ranking tasks, and also allows anyone to easily follow and contribute. We believe that our benchmark could not only reduce the redundant efforts of researchers to re-implement or re-run existing baselines, but also drive more solid and reproducible research on recommender systems.

Recommender systems aim to accurately and actively provide users with potentially interesting items (products, information or services). Deep reinforcement learning has been successfully applied to recommender systems, but still heavily suffer from data sparsity and cold-start in real-world tasks. In this work, we propose an effective way to address such issues by leveraging the pervasive social networks among users in the estimation of action-values (Q). Specifically, we develop a Social Attentive Deep Q-network (SADQN) to approximate the optimal action-value function based on the preferences of both individual users and social neighbors, by successfully utilizing a social attention layer to model the influence between them. Further, we propose an enhanced variant of SADQN, termed SADQN++, to model the complicated and diverse trade-offs between personal preferences and social influence for all involved users, making the agent more powerful and flexible in learning the optimal policies. The experimental results on real-world datasets demonstrate that the proposed SADQNs remarkably outperform the state-of-the-art deep reinforcement learning agents, with reasonable computation cost.

Collaborative filtering (CF) is a widely studied research topic in recommender systems. The learning of a CF model generally depends on three major components, namely interaction encoder, loss function, and negative sampling. While many existing studies focus on the design of more powerful interaction encoders, the impacts of loss functions and negative sampling ratios have not yet been well explored. In this work, we show that the choice of loss function as well as negative sampling ratio is equivalently important. More specifically, we propose the cosine contrastive loss (CCL) and further incorporate it to a simple unified CF model, dubbed SimpleX. Extensive experiments have been conducted on 10 benchmark datasets and compared with 28 existing CF models in total. Surprisingly, the results show that, under our CCL loss and a large negative sampling ratio, SimpleX can surpass most sophisticated state-of-the-art models by a large margin (e.g., max 48.5% improvement in [email protected] over LightGCN). We believe that SimpleX could not only serve as a simple strong baseline to foster future research on CF, but also shed light on the potential research direction towards improving loss function and negative sampling.

Top-N recommendation, which aims to learn user ranking-based preference, has long been a fundamental problem in a wide range of applications. Traditional models usually motivate themselves by designing complex or tailored architectures based on different assumptions. However, the training data of recommender system can be extremely sparse and imbalanced, which poses great challenges for boosting the recommendation performance. To alleviate this problem, in this paper, we propose to reformulate the recommendation task within the causal inference framework, which enables us to counterfactually simulate user ranking-based preferences to handle the data scarce problem. The core of our model lies in the counterfactual question: "what would be the user's decision if the recommended items had been different?''. To answer this question, we firstly formulate the recommendation process with a series of structural equation models (SEMs), whose parameters are optimized based on the observed data. Then, we actively indicate many recommendation lists (called intervention in the causal inference terminology) which are not recorded in the dataset, and simulate user feedback according to the learned SEMs for generating new training samples. Instead of randomly intervening on the recommendation list, we design a learning-based method to discover more informative training samples. Considering that the learned SEMs can be not perfect, we, at last, theoretically analyze the relation between the number of generated samples and the model prediction error, based on which a heuristic method is designed to control the negative effect brought by the prediction error. Extensive experiments are conducted based on both synthetic and real-world datasets to demonstrate the effectiveness of our framework.

Graph convolutional neural networks (GCN) have been the model of choice for graph representation learning, which is mainly due to the effective design of graph convolution that computes the representation of a node by aggregating those of its neighbors. However, existing GCN variants commonly use 1-D graph convolution that solely operates on the object link graph without exploring informative relational information among object attributes. This significantly limits their modeling capability and may lead to inferior performance on noisy and sparse real-world networks. In this paper, we explore 2-D graph convolution to jointly model object links and attribute relations for graph representation learning. Specifically, we propose a computationally efficient dimensionwise separable 2-D graph convolution (DSGC) for filtering node features. Theoretically, we show that DSGC can reduce intra-class variance of node features on both the object dimension and the attribute dimension to learn more effective representations. Empirically, we demonstrate that by modeling attribute relations, DSGC achieves significant performance gain over state-of-the-art methods for node classification and clustering on a variety of real-world networks. The source code for reproducing the experimental results is available at https://github.com/liqimai/DSGC.

Interactive recommendation has drawn widespread attention from both academia and industry due to its effectiveness in real-world mobile applications. Instead of receiving message passively, customers can exploit further with less effort through generated queries. Usually, such systems mainly contain two main components: query generation and item recommendation. In this paper, we propose a novel framework that models both queries and items in shared latent embedding space via a dual sequence transformer structure, which captures customer's potential interest from the prospect of reconciling the historical queries and corresponding customers interactions. We propose a click-through-rate model to generate query candidates, and a session search model for further more precise information. Comprehensive offline and online experiments are conducted, and the results demonstrate that our proposed dual-sequence-transformer based model can better utilize interaction and improve the accuracy of recommendations.

Network embedding is a highly effective method to learn low-dimensional node vector representations with original network structures being well preserved. However, existing network embedding algorithms are mostly developed for a single network, which fails to learn generalized feature representations across different networks. In this article, we study a cross-network node classification problem, which aims at leveraging the abundant labeled information from a source network to help classify the unlabeled nodes in a target network. To succeed in such a task, transferable features should be learned for nodes across different networks. To this end, a novel cross-network deep network embedding (CDNE) model is proposed to incorporate domain adaptation into deep network embedding in order to learn label-discriminative and network-invariant node vector representations. On the one hand, CDNE leverages network structures to capture the proximities between nodes within a network, by mapping more strongly connected nodes to have more similar latent vector representations. On the other hand, node attributes and labels are leveraged to capture the proximities between nodes across different networks by making the same labeled nodes across networks have aligned latent vector representations. Extensive experiments have been conducted, demonstrating that the proposed CDNE model significantly outperforms the state-of-the-art network embedding algorithms in cross-network node classification.

In this paper, the task of cross-network node classification, which leverages the abundant labeled nodes from a source network to help classify unlabeled nodes in a target network, is studied. The existing domain adaptation algorithms generally fail to model the network structural information, and the current network embedding models mainly focus on single-network applications. Thus, both of them cannot be directly applied to solve the cross-network node classification problem. This motivates us to propose an adversarial cross-network deep network embedding (ACDNE) model to integrate adversarial domain adaptation with deep network embedding so as to learn network-invariant node representations that can also well preserve the network structural information. In ACDNE, the deep network embedding module utilizes two feature extractors to jointly preserve attributed affinity and topological proximities between nodes. In addition, a node classifier is incorporated to make node representations label-discriminative. Moreover, an adversarial domain adaptation technique is employed to make node representations network-invariant. Extensive experimental results demonstrate that the proposed ACDNE model achieves the state-of-the-art performance in cross-network node classification.

Network Embedding (NE) aims at modeling network graph by encoding vertices and edges into a low-dimensional space. These learned vectors which preserve proximities can be used for subsequent applications, such as vertex classification and link prediction. Skip-gram with negative sampling is the most widely used method for existing NE models to approximate their objective functions. However, this method only focuses on learning representation from the local connectivity of vertices (i.e., neighbors). In real-world scenarios, a vertex may have multifaceted aspects and should belong to overlapping communities. For example, in a social network, a user may subscribe to political, economic and sports channels simultaneously, but the politics share more common attributes with the economy and less with the sports. In this paper, we propose an adversarial learning approach for modeling overlapping communities of vertices. Each community and vertex are mapped into an embedding space, while we also learn the association between each pair of community and vertex. The experimental results show that our proposed model not only can outperform the state-of-the-art (including GANs-based) models on vertex classification tasks but also can achieve superior performances on overlapping community detection.

Conversion rate (CVR) prediction is becoming increasingly important in the multi-billion dollar online display advertising industry. It has two major challenges: firstly, the scarce user history data is very complicated and non-linear; secondly, the time delay between the clicks and the corresponding conversions can be very large, e.g., ranging from seconds to weeks. Existing models usually suffer from such scarce and delayed conversion behaviors. In this paper, we propose a novel deep learning framework to tackle the two challenges. Specifically, we extract the pre-trained embedding from impressions/clicks to assist in conversion models and propose an inner/self-attention mechanism to capture the fine-grained personalized product purchase interests from the sequential click data. Besides, to overcome the time-delay issue, we calibrate the delay model by learning dynamic hazard function with the abundant post-click data more in line with the real distribution. Empirical experiments with real-world user behavior data prove the effectiveness of the proposed method.

In this article, we study solder paste inspection (SPI), an important stage that is used in the semiconductor manufacturing industry, where abnormal boards should be detected. A highly accurate SPI can substantially reduce human expert involvement, as well as reduce the waste in disposing of the boards in good condition. A key difference today is that because of increasing demand in board customization, the number of board types increases substantially and quantity of the boards produced in each type decreases. Thus, the previous approaches where a fine-tuned model is developed for each board type are no longer viable. Intrinsically, our problem is an anomaly detection problem. A major specialty in today’s SPI is that the target tasks for prediction cannot be fully pre-determined due to context changes during the solder paste printing stage. Our experiences show that a conventional approach to first define a set of tasks and train these tasks offline will lead to low accuracy. Here, we propose a novel multi-task approach, where the performance of all target tasks is ensured simultaneously. We note that the SPI process is streamlined and automatic, allowing the SPI time for only a few seconds. We propose a fast clustering algorithm that reuses existing models to avoid retraining and fine tune in the inference phase. We evaluate our approach using 3-month data collected from production lines. We show that we can reduce 81.28% of false alarms. This can translate to annual savings of $11.3 million.

Improving the performance of click-through rate (CTR) prediction remains one of the core tasks in online advertising systems. With the rise of deep learning, CTR prediction models with deep networks remarkably enhance model capacities. In deep CTR models, exploiting users' historical data is essential for learning users' behaviors and interests. As existing CTR prediction works neglect the importance of the temporal signals when embed users' historical clicking records, we propose a time-aware attention model which explicitly uses absolute temporal signals for expressing the users' periodic behaviors and relative temporal signals for expressing the temporal relation between items. Besides, we propose a regularized adversarial sampling strategy for negative sampling which eases the classification imbalance of CTR data and can make use of the strong guidance provided by the observed negative CTR samples. The adversarial sampling strategy significantly improves the training efficiency, and can be co-trained with the time-aware attention model seamlessly. Experiments are conducted on real-world CTR datasets from both in-station and out-station advertising places.

Task Relation Discovery (TRD), i.e., reveal the relation of tasks, has notable value: it is the key concept underlying Multi-task Learning (MTL) and provides a principled way for identifying redundancies across tasks. However, task relation is usually specifically determined by data scientist resulting in the additional human effort for TRD, while transfer based on brute-force methods or mere training samples may cause negative effects which degrade the learning performance. To avoid negative transfer in an automatic manner, our idea is to leverage commonly available context attributes in nowadays systems, i.e., the metadata. In this paper, we, for the first time, introduce metadata into TRD for MTL and propose a novel Metadata Clustering method, which jointly uses historical samples and additional metadata to automatically exploit the true relatedness. It also avoids the negative transfer by identifying reusable samples between related tasks. Experimental results on five real-world datasets demonstrate that the proposed method is effective for MTL with TRD, and particularly useful in complicated systems with diverse metadata but insufficient data samples. In general, this study helps in automatic relation discovery among partially related tasks and sheds new light on the development of TRD in MTL through the use of metadata as apriori information.

Network Embedding is the task of learning continuous node representations for networks, which has been shown effective in a variety of tasks such as link prediction and node classification. Most of existing works aim to preserve different network structures and properties in low-dimensional embedding vectors, while neglecting the existence of noisy information in many real-world networks and the overfitting issue in the embedding learning process. Most recently, generative adversarial networks (GANs) based regularization methods are exploited to regularize embedding learning process, which can encourage a global smoothness of embedding vectors. These methods have very complicated architecture and suffer from the well-recognized non-convergence problem of GANs. In this paper, we aim to introduce a more succinct and effective local regularization method, namely adversarial training, to network embedding so as to achieve model robustness and better generalization performance. Firstly, the adversarial training method is applied by defining adversarial perturbations in the embedding space with an adaptive L2 norm constraint that depends on the connectivity pattern of node pairs. Though effective as a regularizer, it suffers from the interpretability issue which may hinder its application in certain real-world scenarios. To improve this strategy, we further propose an interpretable adversarial training method by enforcing the reconstruction of the adversarial examples in the discrete graph domain. These two regularization methods can be applied to many existing embedding models, and we take DeepWalk as the base model for illustration in the paper. Empirical evaluations in both link prediction and node classification demonstrate the effectiveness of the proposed methods.

Learning low-dimensional representations of networks has proved effective in a variety of tasks such as node classification, link prediction and network visualization. Existing methods can effectively encode different structural properties into the representations, such as neighborhood connectivity patterns, global structural role similarities and other high-order proximities. However, except for objectives to capture network structural properties, most of them suffer from lack of additional constraints for enhancing the robustness of representations. In this paper, we aim to exploit the strengths of generative adversarial networks in capturing latent features, and investigate its contribution in learning stable and robust graph representations. Specifically, we propose an Adversarial Network Embedding (ANE) framework, which leverages the adversarial learning principle to regularize the representation learning. It consists of two components, i.e., a structure preserving component and an adversarial learning component. The former component aims to capture network structural properties, while the latter contributes to learning robust representations by matching the posterior distribution of the latent representations to given priors. As shown by the empirical results, our method is competitive with or superior to state-of-the-art approaches on benchmark network embedding tasks.