Low-Rank Adaptation (LoRA) is widely used for adapting large language models (LLMs) to specific domains due to its efficiency and modularity. However, vanilla LoRA struggles with task conflicts in multi-task scenarios. Recent works adopt Mixture of Experts (MoE) by treating each LoRA module as an expert, thereby mitigating task interference through multiple specialized LoRA modules. While effective, these methods often isolate knowledge within individual tasks, failing to fully exploit the shared knowledge across related tasks. In this paper, we establish a connection between single LoRA and multi-LoRA MoE, integrating them into a unified framework. We demonstrate that the dynamic routing of multiple LoRAs is functionally equivalent to rank partitioning and block-level activation within a single LoRA. To systematically study the role of expert granularity in multi-task learning, we conduct an in-depth investigation within our unified framework. Our empirical results show that a finer-grained expert partitioning not only yields significant performance gains but also captures more diverse parameter patterns. These empirical findings are supported by our theoretical analysis, which proves that finer granularity expands parameter space diversity and tightens the model's error bound. Building on these findings, we propose Single-ranked Mixture of Experts LoRA (SMoRA ), which embeds MoE into LoRA by treating each rank as an independent expert. With a dynamic rank-wise activation mechanism, SMoRA facilitates a flexible composition of knowledge, enabling the model to learn deeper and more diverse features while mitigating task conflicts. Experiments demonstrate that SMoRA activates fewer parameters yet achieves better performance in multi-task scenarios.

Knowledge editing enables multimodal large language models (MLLMs) to efficiently update outdated or incorrect information. However, existing benchmarks primarily emphasize cognitive-level modifications while lacking a focus on deeper meta-cognitive processes. To bridge this gap, we introduce CogEdit, a novel benchmark designed to evaluate MLLMs' meta-cognitive knowledge editing abilities across three levels: (1) Counterfactual-Driven Editing, assessing self-awareness of knowledge correctness changes; (2) Boundary Constraint Editing, ensuring appropriate generalization without unintended interference; and (3) Noise-Robust Editing, promoting reflective evaluation of uncertain information. To advance meta-cognitive editing, we propose MIND (Meta-cognitive INtegrated Dynamic Knowledge Editing), a framework that constructs a meta-knowledge memory for self-awareness, employs game-theoretic interactions to monitor knowledge activation, and incorporates label refinement for noise-robust updates. Extensive experiments show that MIND significantly outperforms existing cognitive editing approaches, achieving strong performance on both traditional and meta-cognitive knowledge editing benchmarks.

Recent advances in large language models (LLMs) have enabled more semantic-aware recommendations through natural language generation. Existing LLM for recommendation (LLM4Rec) methods mostly operate in a System 1-like manner, relying on superficial features to match similar items based on click history, rather than reasoning through deeper behavioral logic. This often leads to superficial and erroneous recommendations. Motivated by this, we propose ThinkRec, a thinking-based framework that shifts LLM4Rec from System 1 to System 2 (rational system). Technically, ThinkRec introduces a thinking activation mechanism that augments item metadata with keyword summarization and injects synthetic reasoning traces, guiding the model to form interpretable reasoning chains that consist of analyzing interaction histories, identifying user preferences, and making decisions based on target items. On top of this, we propose an instance-wise expert fusion mechanism to reduce the reasoning difficulty. By dynamically assigning weights to expert models based on users' latent features, ThinkRec adapts its reasoning path to individual users, thereby enhancing precision and personalization. Extensive experiments on real-world datasets demonstrate that ThinkRec significantly improves the accuracy and interpretability of recommendations. Our implementations are available at https://github.com/Yu-Qi-hang/ThinkRec.

Nowadays, AI researchers become more and more interested in fine-tuning a pre-trained LLM, whose size has grown to up to over 100B parameters, for their downstream tasks. One approach to fine-tune such huge models is to aggregate device memory from many GPUs. However, this approach introduces prohibitive costs for most data scientists with a limited budget for high-end GPU servers. In this paper, we focus on LLM fine-tuning on a single consumer-grade GPU in a commodity server with limited main memory capacity, which is accessible to most AI researchers. In such a scenario, existing offloading-based methods fail to fine-tune an LLM efficiently due to a lack of holistic intra-server tensor movement management. To this end, we present Ratel, a low-cost, high-performance deep learning training framework that enables efficient 100B-scale model fine-tuning on a commodity server with a consumergrade GPU and limited main memory capacity. The key idea is to add holistic offloading traffic as an optimization dimension for 1) active gradient offloading, and 2) holistic traffic-aware activation swapping mechanism. The experimental results show that 1) Ratel is the first to fine-tune a 175B model on an RTX 4090 and 256 GB main memory, 2) Ratel achieves 2.32x throughput than the state-of-the-art baselines when fine-tuning a small 13B model, and 3) Ratel enables a cheap low-end consumer GPU to have higher cost-effectiveness than a DGX-A100 cluster when fine-tuning a 175B model.

With the wide adoption of GPU and the explosion in data volumes, existing accelerator-centric systems require massive storage access. They adopt high-performance storage devices like NVMe SSDs to scale up single-node systems cost-effectively and leverage the CPU to manage these SSDs. However, they suffer from performance bottlenecks because of the high CPU OS kernel overhead and the CPU memory intermediated data transfer. To address this issue, GPU-initiated and GPU-managed SSD management is proposed to allow the GPU to fully manipulate SSDs: 1) direct data transfer from SSD to GPU memory (data plane) and 2) GPU-managed SSD control (control plane). This can potentially enable these GPU systems to fully leverage the SSD bandwidth. However, we still identify two severe issues. First, the GPU-management SSD control leads to low GPU Streaming Multiprocessor utilization. Second, it leads to the serial execution of SSD accesses with GPU computation, which slows down the overall computing task. To this end, we propose CAM, the first asynchronous GPU-initialized, CPU-managed SSD management for batching storage access. It 1) offloads the SSD control plane from GPU to CPU, thus maximizing GPU streaming multiprocessor utilization, and 2) adopts asynchronous user-friendly APIs that allow programmers to easily overlap GPU computation and SSD I/O operations while keeping a synchronous programming experience. As such, CAM enables us to achieve the best of two worlds: high performance and high programmability. The experimental results show that CAM can perform GNN model training, mergesort, and GEMM up to$\mathbf{1.84}\times, \mathbf{1.5}\times$, and$\mathbf{1.84}\times \mathbf{faster}$compared to the existing state-of-the-art GPU systems, while keeping high programmability.

SSDs are traditionally regarded as a cheap but slow way to scale up GNN training. Several GNN systems explore cheap single-machine single-GPU out-of-core training but fall short in terms of TPC (throughput per monetary cost). The underlying reason is that the existing systems 1) overly focus on minimizing the number of SSD accesses, which results in substantial unnecessary overhead on the CPU side, or 2) exhaust all GPU parallelism to saturate SSD but fail to overlap SSD accesses with GNN computation. In this work, we present Hyperion, a cost-efficient system for terabyte-scale GNN training. We argue that co-optimizing GPU-initiated asynchronous SSD access and GNN computation pipeline enables us to only add cheap NVMe SSDs, rather than expensive GPU servers, to achieve in- memory-like throughput and thus maximal TPC of GNN training. However, this is non-trivial due to imbalanced workloads and interference among IO submission, IO completion, and cache lookup. To tackle the challenges, Hyperion proposes three key designs. First, Hyperion proposes the first GPU-initiated pipeline- friendly asynchronous disk IO stack, which only requires about 1% GPU cores to saturate SSD throughput and wastes no GPU cores between IO submission and completion to fully overlap disk IO and computation. Second, we propose a new GPU-managed, disaggregated, and unified cache that disaggregates cache lookup from disk IO and fully utilizes CPU/GPU memory hierarchy by a unified static cache policy. Third, we propose a GNN-aware general TPC-analytical model that precisely predicts TPC under diverse hardware settings and GNN models and provide a hint to guide users to select hardware, e.g., number of SSDs, under a limited budget to maximize TPC. Experiments demonstrate that Hyperion can improve the TPC by over 3.1x on terabyte-scale graphs compared to SOTA out-of-core baselines and improve 60 x TPC compared to distributed in-memory baselines.

Large Language Models (LLMs) for Recommendation (LLM4Rec) is a promising research direction that has demonstrated exceptional performance in this field. However, its inability to capture real-time user preferences greatly limits the practical application of LLM4Rec because (i) LLMs are costly to train and infer frequently, and (ii) LLMs struggle to access real-time data (its large number of parameters poses an obstacle to deployment on devices). Fortunately, small recommendation models (SRMs) can effectively supplement these shortcomings of LLM4Rec diagrams by consuming minimal resources for frequent training and inference, and by conveniently accessing real-time data on devices.

Foundation models (FMs), known for their broad cognitive capabilities but often constrained to cloud deployment, and specialized models (SMs), characterized by their lightweight, goal-oriented nature suitable for devices, offer complementary strengths. Traditional cloud-centric paradigms face limitations in real-time performance, personalization, cost, and privacy, highlighting the need for innovative approaches that leverage device-level capabilities. This workshop served as a platform to discuss the rapid advancements and emerging research directions in FM-SM collaboration and co-evolution. Key focus areas included: (i) novel collaborative frameworks bridging cloud FMs and device SMs, (ii) mechanisms for model evolution, knowledge transfer, aggregation, and generation, (iii) integration of multimodal perspectives, particularly for multimedia retrieval tasks relevant to ICMR, (iv) strategies for enhancing robustness, interpretability, and fairness, and (v) the development of new benchmarks and resources. Featuring keynote presentations and peer-reviewed papers on topics ranging from multimodal understanding and reasoning to efficient on-device fine-tuning and mobile agents, the workshop fostered interdisciplinary dialogue.

Information Extraction (IE) and Text Classification (CLS) serve as the fundamental pillars of NLU, with both disciplines relying on analyzing input sequences to categorize outputs into pre-established schemas. However, there is no existing encoder-based model that can unify IE and CLS tasks from this perspective. To fully explore the foundation shared within NLU tasks, we have proposed arecursive method with explicit schema instructor for universal NLU. Specifically, we firstly redefine the true universal information extraction (UIE) with a formal formulation that covers almost all extraction schemas, including quadruples and quintuples which remain unsolved for previous UIE models. Then, we expands the formulation to all CLS and multi-modal NLU tasks. Based on that, we introduce RexUniNLU, an universal NLU solution that employs explicit schema constraints for IE and CLS, which encompasses all IE and CLS tasks and prevent incorrect connections between schema and input sequence. To avoid interference between different schemas, we reset the position ids and attention mask matrices. Extensive experiments are conducted on IE, CLS in both English and Chinese, and multi-modality, revealing the effectiveness and superiority. Our codes are publicly released athttps://modelscope.cn/models/iic/nlp_deberta_rex-uninlu_chinese-base.

Treatment effect estimation from observational data is a fundamental problem in causal inference, and its critical challenge is to address the confounding bias arising from the confounders. The effectiveness of the conventional methods proposed to solve this problem depends on the unconfoundedness assumption. In practice, however, the unconfoundedness assumption is frequently violated since we cannot guarantee that all the confounders are measured. To this end, recent studies suggest using auxiliary network architectures to mine information about unmeasured confounders in the data to relax this assumption. However, these methods cannot address the confounding bias from unmeasured confounders unrelated to the network information. Inspired by the insight that some neighboring features that influence one's treatment choice (e.g., which movie to watch) but do not affect the outcome (e.g., assessment of the movie) can be treated as instrumental variables (IVs), we propose a novel Network Instrumental Variable Regression (NetIV) framework exploits IV information from neighborhoods to perform a two-stage regression for treatment effect estimation. Extensive experiments demonstrate that our NetIV method outperforms the state-of-the-art methods for treatment effect estimation in the presence of unmeasured confounders.

The sparse portfolio optimization (SPO) problem is increasingly crucial in portfolio management, focusing on selecting a few stocks with the potential for strong market performance. However, sparse portfolio strategies often face significant short-term drawdowns during periods of market volatility. To this end, a news-driven portfolio strategy offers valuable insights to capture sudden market changes. Nevertheless, it encounters two main challenges:how to reasonably map the relationships between news and stocksandhow to effectively utilize the irregular timing of news releases. To tackle the SPO problem in fluctuating markets while addressing these challenges, we propose a novel news-driven sparse portfolio strategy, named SPIN. Specifically, SPIN not only leverages industry-specific group structures existing among stocks for a more reasonable news-stock mapping and models news sequential patterns based on our devised novel news-driven forecaster to handle the irregularity of news releases. We rigorously prove that SPIN achieves a sub-linear regret. Extensive experiments on three real-world datasets demonstrate SPIN's superiority over state-of-the-art portfolio strategies in terms of cumulative wealth and short-term drawdowns.

Recommender systems primarily utilize user–item interactions (i.e., collaborative information) and auxiliary textual information (i.e., semantic information) to infer user preferences and provide recommendations. With the advancement in large language models (LLMs), attempts have been made to incorporate their remarkable language comprehension capabilities into recommendation tasks. However, existing LLM4Rec methods face challenges in seamlessly integrating both collaborative and semantic information, as there is an inherent gap between these two types of data. Moreover, these methods struggle to capture the fine-grained distinctions in user preferences, which are essential in recommendation tasks, due to the loss design of LLMs. To address these issues, we propose a multi-stage prompt-tuning method for leveraging pre-trained LLMs in various recommendation tasks, named SCRec. Specifically, SCRec leverages Semantic and Collaborative information as supervision signals in two distinct stages: the semantic prompt-tuning stage and the collaborative prompt-tuning stage. This method breaks down user and item representations into semantic and collaborative perspectives, enabling a pre-trained LLM to first deduce the qualitative preferences of users over items from semantic information, and then generate quantitative recommendations from collaborative information. In addition, we propose a meta-mapping approach to provide personalized mapping functions for encoding collaborative information and integrate a novel numeric-informed head based on MSE loss for LLM in the second stage, which helps to better capture fine-grained distinctions in user preferences. Experiments on three public datasets for rating prediction and top-N recommendation tasks demonstrate that our method surpasses both conventional and LLM-based techniques, showing the strength of sequentially merging semantic and collaborative information in recommendation tasks.

Knowledge tracing (KT) is a fundamental task in intelligent education aimed at tracking students’ knowledge status and predicting their performance on new questions. The primary challenge in KT is accurately inferring a high-quality representation of students’ knowledge state that effectively captures their understanding of questions. However, existing methods are typically developed under the assumption that students’ behaviors directly reflect their knowledge state, which may not hold true especially in online learning scenarios. Abnormal behaviors exhibited by students, such as guessing and plagiarism, can introduce biases into the data, making it difficult to accurately assess students’ true knowledge state. To address this limitation, we propose a novel DebiAsed Cognition rEpresentation (DACE) modeling approach. This approach introduces a novel adversarial training strategy based on information bottleneck theory to obtain a debiased knowledge state representation that retains only the most reliable information for accurately predicting students’ performance on new questions. Moreover, we design a novel contrastive learning module through embedding-based augmentation to further enhance the robustness and generalizability of the learned knowledge state representation. We conduct extensive experiments on three public KT datasets and the newly released dataset BaiPy to demonstrate the superiority of our model over strong baselines, particularly when confronted with biased data. Our code and datasets are available at https://github.com/lvXiangwei/DACE.git .

Estimating individual treatment effects in networked observational data is a crucial and increasingly recognized problem. One major challenge of this problem is violating the stable unit treatment value assumption (SUTVA), which posits that a unit’s outcome is independent of others’ treatment assignments. However, in network data, a unit’s outcome is influenced not only by its treatment (i.e., direct effect) but also by the treatments of others (i.e., spillover effect) since the presence of interference. Moreover, the interference from other units is always heterogeneous (e.g., friends with similar interests have a different influence than those with different interests). In this article, we focus on the problem of estimating individual treatment effects (including direct effect and spillover effect) under heterogeneous interference in networks. To address this problem, we propose a novel dual weighting regression (DWR) algorithm by simultaneously learning attention weights to capture the heterogeneous interference from neighbors and sample weights to eliminate the complex confounding bias in networks. We formulate the learning process as a bi-level optimization problem. Theoretically, we give a generalization error bound for the expected estimation error of the individual treatment effects. Extensive experiments on four benchmark datasets demonstrate that the proposed DWR algorithm outperforms the state-of-the-art methods in estimating individual treatment effects under heterogeneous network interference.

In the training procedures of many real-world learning models, gathering and annotating decent amounts of labeled data can be cost-prohibitive. To mitigate this data-hungry problem, active learning (AL) and semi-supervised learning (SSL) are frequently adopted as two effective but often isolated means. Some recent studies explored the potential of combining AL and SSL to better probe the unlabeled data. However, almost all these contemporary SSL-AL works use a simple combination strategy, ignoring SSL and AL's inherent relation. Further, other methods suffer from high computational costs when dealing with large-scale, high-dimensional datasets. Motivated by the industry practice of labeling data, we first propose an innovativeInconsistency-based virtual aDvErsarialActiveLearning (IDEAL) algorithm to further investigate SSL-AL's potential superiority and achieve mutual enhancement of AL and SSL, i.e., SSL propagates label information to unlabeled samples and provides smoothed embeddings for AL, while AL excludes samples with inconsistent predictions and considerable uncertainty for SSL. We estimate unlabeled samples’ inconsistency by augmentation strategies of different granularities, including fine-grained continuous perturbation exploration and coarse-grained data transformations. Moreover, to solve the problem that the utilization efficiency of unlabeled samples is still insufficient in the process of semi-supervised training, we extend our IDEAL to a curriculum-guided version, namely SPL-IDEAL algorithm. The SPL-IDEAL algorithm can regularize the training process towards better regions in parameter space and denoise the pseudo labels with low confidence, achieving better performance. The extensive experiments, in both text and image benchmark datasets, validate the effectiveness of our proposed IDEAL and SPL-IDEAL algorithms, comparing them against state-of-the-art baselines. Two real-world case studies visualize the practical industrial value of applying and deploying the proposed data sampling algorithms.

Driven by empirical risk minimization, machine learning algorithm tends to exploit subtle statistical correlations existing in the training environment for prediction, while the spurious correlations are unstable across environments, leading to poor generalization performance. Accordingly, the problem of the Out-of-distribution (OOD) generalization aims to exploit an invariant/stable relationship between features and outcomes that generalizes well on all possible environments. To address the spurious correlation induced by the selection bias, in this article, we propose a novel Clique-based Causal Feature Separation (CCFS) algorithm by explicitly incorporating the causal structure to identify causal features of outcome for OOD generalization. Specifically, the proposed CCFS algorithm identifies the largest clique in the learned causal skeleton. Theoretically, we guarantee that either the largest clique or the rest of the causal skeleton is exactly the set of all causal features of the outcome. Finally, we separate the causal features from the non-causal ones with a sample-reweighting decorrelator for OOD prediction. Extensive experiments validate the effectiveness of the proposed CCFS method on both causal feature identification and OOD generalization tasks.

Recommendation performance usually exhibits a long-tail distribution over users — a small portion of head users enjoy much more accurate recommendation services than the others. We reveal two sources of this performance heterogeneity problem: the uneven distribution of historical interactions (a natural source); and the biased training of recommender models (a model source). As addressing this problem cannot sacrifice the overall performance, a wise choice is to eliminate the model bias while maintaining the natural heterogeneity. The key to debiased training lies in eliminating the effect of confounders that influence both the user's historical behaviors and the next behavior. The emerging causal recommendation methods achieve this by modeling the causal effect between user behaviors, however potentially neglect unobserved confounders (e.g., friend suggestions) that are hard to measure in practice. To address unobserved confounders, we resort to the front-door adjustment (FDA) in causal theory and propose a causal multi-teacher distillation framework (CausalD). FDA requires proper mediators in order to estimate the causal effects of historical behaviors on the next behavior. To achieve this, we equip CausalD with multiple heterogeneous recommendation models to model the mediator distribution. Then, the causal effect estimated by FDA is the expectation of recommendation prediction over the mediator distribution and the prior distribution of historical behaviors, which is technically achieved by multi-teacher ensemble. To pursue efficient inference, CausalD further distills multiple teachers into one student model to directly infer the causal effect for making recommendations. We instantiate CausalD on two representative models, DeepFM and DIN, and conduct extensive experiments on three real-world datasets, which validate the superiority of CausalD over state-of-the-art methods. Through in-depth analysis, we find that CausalD largely improves the performance of tail users, reduces the performance heterogeneity, and enhances the overall performance.

Dependency-based models are widely used to extract semantic relations in text. Most existing dependency-based models establish stacked structures to merge contextual and dependency information, which encode the contextual information first and then encode the dependency information. However, this unidirectional information flow weakens the representation of words in the sentence, which further restricts the performance of existing models. To establish bidirectional information flow, a dual attention graph convolutional network (DAGCN) with a parallel structure is proposed. Most importantly, DAGCN can build multi-turn interactions between contextual and dependency information to imitate the multi-turn looking-back actions of human beings. In addition, multi-layer adjacency matrix-aware multi-head attention (AMAtt), including context-to-dependency attention and dependency-to-context attention, is carefully designed as a merge mechanism in the parallel structure to preserve the structural information of sentences and dependency trees during interactions. Furthermore, DAGCN is evaluated on the popular PubMed dataset, TACRED dataset and SemEval 2010 Task 8 dataset to demonstrate its validity. Experimental results show that our model outperforms the existing dependency-based models.

In recommender systems, click behaviors play a fundamental role in mining users’ interests and training models (clicked items as positive samples). Such signals are implicit feedback and are arguably less representative of users’ inherent interests. Most existing works denoise implicit feedback by introducing external signals, such as gaze, dwell time, and “like” behaviors. However, such explicit feedback is not always routinely available, or might be problematic to collect on a large scale. In this paper, we identify that an interaction’s related structural patterns in its neighborhood graph are potentially correlated with some outcome of implicit feedback (i.e., users’ ratings after consuming items), analogous to findings in other domains such as social networks. Inspired by this finding, we propose a novel Structure LEarning based Denoising (SLED) framework for denoising recommendation without explicit signals, which consists of two phases: center-aware graph structure learning and denoised recommendation . Phase 1 pre-trains a structural encoder in a self-supervised manner and learns to capture an interaction’s related structural patterns in its neighborhood graph. Phase 2 transfers the structure encoder to downstream recommendation datasets, which helps to down-weight the effect of noisy interactions on user interest modeling and loss calculation. We collect a relatively noisy industrial dataset across several days during a period of product promotion festival. Extensive experiments on this dataset and multiple public datasets demonstrate that the proposed SLED framework can significantly improve the recommendation quality over various base recommendation models.

Tackling the pervasive issue of data sparsity in recommender systems, we present an insightful investigation into the burgeoning area of non-overlapping cross-domain recommendation, a technique that facilitates the transfer of interaction knowledge across domains without necessitating inter-domain user/item correspondence. Existing approaches have predominantly depended on auxiliary information, such as user reviews and item tags, to establish inter-domain connectivity, but these resources may become inaccessible due to privacy and commercial constraints. To address these limitations, our study introduces an in-depth exploration of Target-unknown Cross-domain Recommendation (CDR), which contends with the distinct challenge of lacking target domain information during the training phase in the source domain. We illustrate two critical obstacles inherent to Target-unknown CDR: the lack of an inter-domain bridge due to insufficient user/item correspondence or side information and the potential pitfalls of source-domain training biases when confronting distribution shifts across domains. To surmount these obstacles, we propose the CMCDR framework, a novel approach that leverages causal mechanisms extracted from meta-user/item representations. The CMCDR framework employs a vector-quantized encoder–decoder architecture, enabling the disentanglement of user/item characteristics. We posit that domain-transferable knowledge is more readily discernible from user/item characteristics, i.e., the meta-representations, rather than raw users and items. Capitalizing on these meta-representations, our CMCDR framework adeptly incorporates an attention-driven predictor that approximates the front-door adjustment method grounded in causal theory. This cutting-edge strategy effectively mitigates source-domain training biases and enhances generalization capabilities against distribution shifts. Extensive experiments demonstrate the empirical effectiveness and the rationality of CMCDR for target-unknown cross-domain recommendation.

In recommendation scenarios, there are two long-standing challenges, i.e., selection bias and data sparsity, which lead to a significant drop in prediction accuracy for both Click-Through Rate (CTR) and post-click Conversion Rate (CVR) tasks. To cope with these issues, existing works emphasize on leveraging Multi-Task Learning (MTL) frameworks (Category 1) or causal debiasing frameworks (Category 2) to incorporate more auxiliary data in the entire exposure/inference space $\mathcal{D}$ or debias the selection bias in the click/training space ${\mathcal{O}}$. However, these two kinds of solutions cannot effectively address the not-missing-at-random problem and debias the selection bias in ${\mathcal{O}}$ to fit the inference in $\mathcal{D}$. To fill the research gaps, we propose a Direct entire-space Causal Multi-Task framework, namely DCMT, for post-click conversion prediction in this paper. Specifically, inspired by users’ decision process of conversion, we propose a new counterfactual mechanism to debias the selection bias in $\mathcal{D}$, which can predict the factual CVR and the counterfactual CVR under the soft constraint of a counterfactual prior knowledge. Extensive experiments demonstrate that our DCMT can improve the state-of-the-art methods by an average of 1.07% in term of CVR AUC on the offline datasets and 0.75% in term of PV-CVR on the online A/B test (the Alipay Search). Such improvements can increase millions of conversions per week in real industrial applications, e.g., the Alipay Search.

The NP-hard combinatorial Virtual Network Embedding (VNE) Problem refers to finding the node and edge mapping between a virtual net (request) and the physical net (resource). Learning-based methods are recently devised beyond traditional heuristic solvers. However, the efficiency and scalability hinder its applicability as reinforcement learning (RL) is often adopted in an auto-regressive node-by-node mapping manner to handle complex mapping constraints, for each coming request for mapping. Moreover, existing learning-based works often independently consider each online request, limiting the long-term online service performance. In this paper, we present a synergistic Global-And-Local learning approach for the VNE problem (GAL-VNE). At the global level across requests, RL is employed to capture the cross-request relation for better global resource accommodation to improve overall performance. At the local level within each request, we aim to replace the sequential decision-making procedure which relies much on the network size, with a more efficient one-shot solution generation scheme. The main challenge for such a one-shot model is how to encode the constraints under an end-to-end learning and inference paradigm. Accordingly, within the "rank-then-search" paradigm, we propose to first pretrain a graph neural network (GNN)-based node ranker with imitation supervision from an off-the-shelf solver (moderately expensive yet high quality), which is meanwhile regularized by a neighboring smooth prior. Then RL is used to finetune the GNN ranker whose supervision directly refers to the final (undifferentiable) business objectives concerning revenue and cost, etc. Experiments on benchmarks show that our method outperforms classic and learning-based methods in both efficacy and efficiency.

Domain generalization (DG) is a prevalent problem in real-world applications, which aims to train well-generalized models for unseen target domains by utilizing several source domains. Since domain labels, i.e., which domain each data point is sampled from, naturally exist, most DG algorithms treat them as a kind of supervision information to improve the generalization performance. However, the original domain labels may not be the optimal supervision signal due to the lack of domain heterogeneity, i.e., the diversity among domains. For example, a sample in one domain may be closer to another domain, its original label thus can be the noise to disturb the generalization learning. Although some methods try to solve it by re-dividing domains and applying the newly generated dividing pattern, the pattern they choose may not be the most heterogeneous due to the lack of the metric for heterogeneity. In this paper, we point out that domain heterogeneity mainly lies in variant features under the invariant learning framework. With contrastive learning, we propose a learning potential-guided metric for domain heterogeneity by promoting learning variant features. Then we notice the differences between seeking variance-based heterogeneity and training invariance-based generalizable model. We thus propose a novel method called H eterogeneity-based Two-stage Contrastive Learning (HTCL) for the DG task. In the first stage, we generate the most heterogeneous dividing pattern with our contrastive metric. In the second stage, we employ an invariance-aimed contrastive learning by re-building pairs with the stable relation hinted by domains and classes, which better utilizes generated domain labels for generalization learning. Extensive experiments show HTCL better digs heterogeneity and yields great generalization performance.

Multi-modal Sentiment Analysis (MSA) is a hotspot of multi-modal fusion. To make full use of the correlation and complementarity between modalities in the process of fusing multi-modal data, we propose a two-stage framework of Contrastive Language-Audio Pre-training (CLAP) for the MSA task: 1) Making contrastive pre-training on an unlabeled large-scaled external data to yield better single-modal representations; 2) Adopting a Transformer-based multi-modal fusion module, to achieve further single-modal feature optimization and sentiment prediction via the task-driven training process. Our work fully demonstrates the importance and necessity of core elements such as pre-training, contrastive learning, and representation learning for the MSA task and significantly outperforms existing methods on two well-recognized MSA benchmarks.

In marketing recommendations, the campaign organizers will distribute coupons to users to encourage consumption. In general, a series of strategies are employed to interfere with the coupon distribution process, leading to a growing imbalance between user-coupon interactions, resulting in a bias in the estimation of conversion probabilities. We refer to the estimation bias as the matching bias. In this paper, we explore how to alleviate the matching bias from the causal-effect perspective. We regard the historical distributions of users and coupons over each other as confounders and characterize the matching bias as a confounding effect to reveal and eliminate the spurious correlations between user-coupon representations and conversion probabilities. Then we propose a new training paradigm named De-Matching Bias Recommendation (DMBR) to remove the confounding effects during model training via the backdoor adjustment. We instantiate DMBR on two representative models: DNN and MMOE, and conduct extensive offline and online experiments to demonstrate the effectiveness of our proposed paradigm.

Legal judgment prediction (LJP) is a significant task in legal intelligence, which aims to assist the judges and determine the judgment result based on the case's fact description. The judgment result consists of law articles, charge, and prison term. The law articles serve as the basis for the charge and the prison term, which can be divided into two types, named as charge-related law article and term-related law article, respectively. Recently, many methods have been proposed and made tremendous progress in LJP. However, the existing methods only focus on the prediction of the charge-related law articles, ignoring the term-related law articles (e.g., laws about lenient treatment), which limits the performance in the prison term prediction. In this paper, following the actual legal process, we expand the law article prediction as a multi-label classification task that includes both the charge-related law articles and term-related law articles and propose a novel multi-law aware LJP (ML-LJP) method to improve the performance of LJP. Given the case's fact description, firstly, the label (e.g., law article and charge) definitions in the Code of Law are used to transform the representation of the fact into several label-specific representations and make the prediction of the law articles and the charge. To distinguish the similar content of different label definitions, contrastive learning is conducted in the training. Then, a graph attention network (GAT) is applied to learn the interactions among the multiple law articles for the prediction of the prison term. Since numbers (e.g., amount of theft and weight of drugs) are important for LJP but often ignored by conventional encoders, we design a corresponding number representation method to locate and better represent these effective numbers. Extensive experiments on real-world dataset show that our method achieves the best results compared to the state-of-the-art models, especially in the task of prison term prediction where ML-LJP achieves a 10.07% relative improvement over the best baseline.

In the field of music information retrieval (MIR), cover song identification (CSI) is a challenging task that aims to identify cover versions of a query song from a massive collection. Existing works still suffer from high intra-song variances and inter-song correlations, due to the entangled nature of version-specific and version-invariant factors in their modeling. In this work, we set the goal of disentangling version-specific and version-invariant factors, which could make it easier for the model to learn invariant music representations for unseen query songs. We analyze the CSI task in a disentanglement view with the causal graph technique, and identify the intra-version and inter-version effects biasing the invariant learning. To block these effects, we propose the disentangled music representation learning framework (DisCover) for CSI. DisCover consists of two critical components: (1) Knowledge-guided Disentanglement Module (KDM) and (2) Gradient-based Adversarial Disentanglement Module (GADM), which block intra-version and inter-version biased effects, respectively. KDM minimizes the mutual information between the learned representations and version-variant factors that are identified with prior domain knowledge. GADM identifies version-variant factors by simulating the representation transitions between intra-song versions, and exploits adversarial distillation for effect blocking. Extensive comparisons with best-performing methods and in-depth analysis demonstrate the effectiveness of DisCover and the and necessity of disentanglement for CSI.

Device Model Generalization (DMG) is a practical yet under-investigated research topic for on-device machine learning applications. It aims to improve the generalization ability of pre-trained models when deployed on resource-constrained devices, such as improving the performance of pre-trained cloud models on smart mobiles. While quite a lot of works have investigated the data distribution shift across clouds and devices, most of them focus on model fine-tuning on personalized data for individual devices to facilitate DMG. Despite their promising, these approaches require on-device re-training, which is practically infeasible due to the overfitting problem and high time delay when performing gradient calculation on real-time data. In this paper, we argue that the computational cost brought by fine-tuning can be rather unnecessary. We consequently present a novel perspective to improving DMG without increasing computational cost, i.e., device-specific parameter generation which directly maps data distribution to parameters. Specifically, we propose an efficient Device-cloUd collaborative parametErs generaTion framework (DUET). DUET is deployed on a powerful cloud server that only requires the low cost of forwarding propagation and low time delay of data transmission between the device and the cloud. By doing so, DUET can rehearse the device-specific model weight realizations conditioned on the personalized real-time data for an individual device. Importantly, our DUET elegantly connects the cloud and device as a “duet” collaboration, frees the DMG from fine-tuning, and enables a faster and more accurate DMG paradigm. We conduct an extensive experimental study of DUET on three public datasets, and the experimental results confirm our framework’s effectiveness and generalisability for different DMG tasks.

Session-based Recommendation aims at predicting the next interacted item based on short anonymous behavior sessions. However, existing solutions neglect to model two inherent properties of sequential representing distributions, i.e., hierarchy structures resulted from item popularity and collaborations existing in both intra- and inter-session. Tackling with these two factors at the same time is challenging. On the one hand, traditional Euclidean space utilized in previous studies fails to capture hierarchy structures due to a restricted representation ability. On the other hand, the intuitive apply of hyperbolic geometry could extract hierarchical patterns but more emphasis on degree distribution weakens intra- and inter-session collaborations. To address the challenges, we propose a Hierarchy-Aware Dual Clustering Graph Network (HADCG) model for session-based recommendation. Towards the first challenge, we design the hierarchy-aware graph modeling module which converts sessions into hyperbolic session graphs, adopting hyperbolic geometry in propagation and attention mechanism so as to integrate chronological and hierarchical information. As for the second challenge, we introduce the deep dual clustering module which develops a two-level clustering strategy, i.e., information regularizer for intra-session clustering and contrastive learner for inter-session clustering, to enhance hyperbolic representation learning from collaborative perspectives and further promote recommendation performance. Extensive experiments on three real-world datasets demonstrate the effectiveness of the proposed HADCG.

Domain generalization (DG) aims to learn from multiple source domains a model that can generalize well on unseen target domains. Existing DG methods mainly learn the representations with invariant marginal distribution of the input features, however, the invariance of the conditional distribution of the labels given the input features is more essential for unknown domain prediction. Meanwhile, the existing of unobserved confounders which affect the input features and labels simultaneously cause spurious correlation and hinder the learning of the invariant relationship contained in the conditional distribution. Interestingly, with a causal view on the data generating process, we find that the input features of one domain are valid instrumental variables for other domains. Inspired by this finding, we propose an instrumental variable-driven DG method (IV-DG) by removing the bias of the unobserved confounders with two-stage learning. In the first stage, it learns the conditional distribution of the input features of one domain given input features of another domain. In the second stage, it estimates the relationship by predicting labels with the learned conditional distribution. Theoretical analyses and simulation experiments show that it accurately captures the invariant relationship. Extensive experiments on real-world datasets demonstrate that IV-DG method yields state-of-the-art results.

In this paper, we focus on the problem of stable prediction across unknown test data, where the test distribution might be different from the training one and is always agnostic when model training. In such a case, previous machine learning methods might exploit subtly spurious correlations induced by non-causal variables in training data for prediction. Those spurious correlations are changeable across data, leading to instability of prediction across unknown test data. To address this problem, we propose a conditional independence test based algorithm to screen out part of non-causal features and reduce those spurious correlations for a more stable prediction by leveraging a seed variable. We show, both theoretically and with empirical experiments, that our algorithm can precisely screen out the isolated non-causal variables, which have no causal relationship with other variables, and remove the spurious correlations induced by them, increasing the stability of prediction across unknown test data. Extensive experiments on both synthetic and real-world datasets demonstrate that our algorithm outperforms state-of-the-art methods for stable prediction across unknown test data.

In observational studies, confounder separation and balancing are the fundamental problems of treatment effect estimation. Most of the previous methods focused on addressing the problem of confounder balancing by treating all observed pre-treatment variables as confounders, ignoring confounder separation. In general, not all the observed pre-treatment variables are confounders that refer to the common causes of the treatment and the outcome, some variables only contribute to the treatment (i.e., instrumental variables) and some only contribute to the outcome (i.e., adjustment variables). Balancing those non-confounders, including instrumental variables and adjustment variables, would generate additional bias for treatment effect estimation. By modeling the different causal relations among observed pre-treatment variables, treatment variables and outcome variables, we propose a synergistic learning framework to i) separate confounders by learning decomposed representations of both confounders and non-confounders, ii) balance confounder with sample re-weighting technique, and simultaneously iii) estimate the treatment effect in observational studies via counterfactual inference. Empirical results on synthetic and real-world datasets demonstrate that the proposed method can precisely decompose confounders and achieve a more precise estimation of treatment effect than baselines.

Influenced by the great success of deep learning via cloud computing and the rapid development of edge chips, research in artificial intelligence (AI) has shifted to both of the computing paradigms, i.e., cloud computing and edge computing. In recent years, we have witnessed significant progress in developing more advanced AI models on cloud servers that surpass traditional deep learning models owing to model innovations (e.g., Transformers, Pretrained families), explosion of training data and soaring computing capabilities. However, edge computing, especially edge and cloud collaborative computing, are still in its infancy to announce their success due to the resource-constrained IoT scenarios with very limited algorithms deployed. In this survey, we conduct a systematic review for both cloud and edge AI. Specifically, we are the first to set up the collaborative learning mechanism for cloud and edge modeling with a thorough review of the architectures that enable such mechanism. We also discuss potentials and practical experiences of some on-going advanced edge AI topics including pretraining models, graph neural networks and reinforcement learning. Finally, we discuss the promising directions and challenges in this field.

Domain generalization (DG) aims to learn from multiple known source domains a model that can generalize well to unknown target domains. The existing DG methods usually exploit the fusion of shared multi-source data to train a generalizable model. However, tremendous data is distributed across lots of places nowadays that can not be shared due to privacy policies. In this paper, we tackle the problem of federated domain generalization where the source datasets can only be accessed and learned locally for privacy protection. We propose a novel framework called Collaborative Semantic Aggregation and Calibration (CSAC) to enable this challenging problem. To fully absorb multi-source semantic information while avoiding unsafe data fusion, we conduct data-free semantic aggregation by fusing the models trained on the separated domains layer-by-layer. To address the semantic dislocation problem caused by domain shift, we further design cross-layer semantic calibration with an attention mechanism to align each semantic level and enhance domain invariance. We unify multi-source semantic learning and alignment in a collaborative way by repeating the semantic aggregation and calibration alternately, keeping each dataset localized, and the data privacy is carefully protected. Extensive experiments show the significant performance of our method in addressing this challenging problem.

Learning individualized causal effect (ICE) plays a vital role in various fields of big data analysis, ranging from fine-grained policy evaluation to personalized treatment development. However, the presence of unmeasured confounders increases the difficulty of estimating ICE in real-world scenarios. A wide range of methods have been proposed to address the unmeasured confounders with the aid of instrument variable (IV), which sources from the treatment randomization. The performance of these methods relies on the well-predefined IVs that satisfy the unconfounded instruments assumption (i.e., the IVs are independent with the unmeasured confounders given observed covariates), which is untestable and leads to finding a valid IV becomes an art rather than science. In this paper, we focus on estimating the ICE with confounded instruments that violate the unconfounded instruments assumption. By considering the conditional independence between the set of confounded instruments and the outcome variable, we propose a novel method, named CVAE-IV, to generate a substitute of the unmeasured confounder with a conditional variational autoencoder. Our theoretical analysis guarantees that the generated confounder substitute will identify unbiased ICE. Extensive experiments on bias demand prediction and Mendelian randomization analysis verify the effectiveness of our method.

While annotating decent amounts of data to satisfy sophisticated learning models can be cost-prohibitive for many real-world applications. Active learning (AL) and semi-supervised learning (SSL) are two effective, but often isolated, means to alleviate the data-hungry problem. Some recent studies explored the potential of combining AL and SSL to better probe the unlabeled data. However, almost all these contemporary SSL-AL works use a simple combination strategy, ignoring SSL and AL's inherent relation. Further, other methods suffer from high computational costs when dealing with large-scale, high-dimensional datasets. Motivated by the industry practice of labeling data, we propose an innovative Inconsistency-based virtual aDvErsarial Active Learning (IDEAL) algorithm to further investigate SSL-AL's potential superiority and achieve mutual enhancement of AL and SSL, i.e., SSL propagates label information to unlabeled samples and provides smoothed embeddings for AL, while AL excludes samples with inconsistent predictions and considerable uncertainty for SSL. We estimate unlabeled samples' inconsistency by augmentation strategies of different granularities, including fine-grained continuous perturbation exploration and coarse-grained data transformations. Extensive experiments, in both text and image domains, validate the effectiveness of the proposed algorithm, comparing it against state-of-the-art baselines. Two real-world case studies visualize the practical industrial value of applying and deploying the proposed data sampling algorithm.

Waterfall Recommender System (RS), a popular form of RS in mobile applications, is a stream of recommended items consisting of successive pages that can be browsed by scrolling. In waterfall RS, when a user finishes browsing a page, the edge (e.g., mobile phones) would send a request to the cloud server to get a new page of recommendations, known as the paging request mechanism. RSs typically put a large number of items into one page to reduce excessive resource consumption from numerous paging requests, which, however, would diminish the RSs' ability to timely renew the recommendations according to users' real-time interest and lead to a poor user experience. Intuitively, inserting additional requests inside pages to update the recommendations with a higher frequency can alleviate the problem. However, previous attempts, including only non-adaptive strategies (e.g., insert requests uniformly), would eventually lead to resource overconsumption. To this end, we envision a new learning task of edge intelligence named Intelligent Request Strategy Design (IRSD). It aims to improve the effectiveness of waterfall RSs by determining the appropriate occasions of request insertion based on users' real-time intention. Moreover, we propose a new paradigm of adaptive request insertion strategy named Uplift-based On-edge Smart Request Framework (AdaRequest). AdaRequest 1) captures the dynamic change of users' intentions by matching their real-time behaviors with their historical interests based on attention-based neural networks. 2) estimates the counterfactual uplift of user purchase brought by an inserted request based on causal inference. 3) determines the final request insertion strategy by maximizing the utility function under online resource constraints. We conduct extensive experiments on both offline dataset and online A/B test to verify the effectiveness of AdaRequest. Remarkably, AdaRequest has been deployed on the Waterfall RS of Taobao and brought over 3% lift on Gross Merchandise Value (GMV).

In recent years, online short videos have become more popular, especially as an online advertising intermediary. To better understand their effects as advertisements, it is essential to analyze the causal relations of online short videos on consumer behaviors. Our study is based on fine-grained consumer behavior data from a world-leading e-commerce platform, i.e., Taobao.com. We first decompose the total causal effects into informative effects and persuasive effects following a common practice in the economic literature. Moreover, we extract the subjectivity scores of short videos through a dictionary-based subjectivity analysis model and evaluate the correlation between the subjectivity scores and each causal effect. The findings of this paper are as follows: First, both causal effects (i.e., informative and persuasive effects) are significant. Second, these effects have a strong correlation with the short videos' subjectivity scores. Third, the signs of these correlations vary with the prices of the products. Our results not only shed light on the research of how short videos exert influence on online consumers, but also give sellers advice on better video design and recommendation.

Effectively representing users lie at the core of modern recommender systems. Since users’ interests naturally exhibit multiple aspects, it is of increasing interest to develop multi-interest frameworks for recommendation, rather than represent each user with an overall embedding. Despite their effectiveness, existing methods solely exploit the encoder (the forward flow) to represent multiple aspects of interests. However, without explicit regularization, the interest embeddings may not be distinct from each other nor semantically reflect representative historical items. Towards this end, we propose the Re4 framework, which leverages the backward flow to reexamine each interest embedding. Specifically, Re4 encapsulates three backward flows, i.e., 1) Re-contrast, which drives each interest embedding to be distinct from other interests using contrastive learning; 2) Re-attend, which ensures the interest-item correlation estimation in the forward flow to be consistent with the criterion used in final recommendation; and 3) Re-construct, which ensures that each interest embedding can semantically reflect the information of representative items that relate to the corresponding interest. We demonstrate the novel forward-backward multi-interest paradigm on ComiRec, and perform extensive experiments on three real-world datasets. Empirical studies validate that Re4 helps to learn learning distinct and effective multi-interest representations.

In data mining and machine learning, it is commonly assumed that training and test data share the same population distribution. However, this assumption is often violated in practice because of the sample selection bias, which might induce the distribution shift from training data to test data. Such a model-agnostic distribution shift usually leads to prediction instability across unknown test data. This article proposes a novel balance-subsampled stable prediction (BSSP) algorithm based on the theory of fractional factorial design. It isolates the clear effect of each predictor from the confounding variables. A design-theoretic analysis shows that the proposed method can reduce the confounding effects among predictors induced by the distribution shift, improving both the accuracy of parameter estimation and the stability of prediction across unknown test data. Numerical experiments on synthetic and real-world datasets demonstrate that our BSSP algorithm can significantly outperform the baseline methods for stable prediction across unknown test data.

Instrumental variables (IVs), sources of treatment randomization that are conditionally independent of the outcome, play an important role in causal inference with unobserved confounders. However, the existing IV-based counterfactual prediction methods need well-predefined IVs, while it’s an art rather than science to find valid IVs in many real-world scenes. Moreover, the predefined hand-made IVs could be weak or erroneous by violating the conditions of valid IVs. These thorny facts hinder the application of the IV-based counterfactual prediction methods. In this article, we propose a novel Automatic Instrumental Variable decomposition (AutoIV) algorithm to automatically generate representations serving the role of IVs from observed variables (IV candidates). Specifically, we let the learned IV representations satisfy the relevance condition with the treatment and exclusion condition with the outcome via mutual information maximization and minimization constraints, respectively. We also learn confounder representations by encouraging them to be relevant to both the treatment and the outcome. The IV and confounder representations compete for the information with their constraints in an adversarial game, which allows us to get valid IV representations for IV-based counterfactual prediction. Extensive experiments demonstrate that our method generates valid IV representations for accurate IV-based counterfactual prediction.

Causal Inference plays an important role in decision making in many fields, such as social marketing, healthcare, and public policy. One fundamental problem in causal inference is the treatment effect estimation in observational studies when variables are confounded. Controlling for confounding effects is generally handled by propensity score. But it treats all observed variables as confounders and ignores the adjustment variables, which have no influence on treatment but are predictive of the outcome. Recently, it has been demonstrated that the adjustment variables are effective in reducing the variance of the estimated treatment effect. However, how to automatically separate the confounders and adjustment variables in observational studies is still an open problem, especially in the scenarios of high dimensional variables, which are common in the big data era. In this paper, we first propose a Data-Driven Variable Decomposition (D$^2$VD) algorithm, which can 1) automatically separate confounders and adjustment variables with a data-driven approach, and 2) simultaneously estimate treatment effect in observational studies with high dimensional variables. Under standard assumptions, we theoretically prove that our D$^2$VD algorithm can unbiased estimate treatment effect and achieve lower variance than traditional propensity score based methods. Moreover, to address the challenges from high-dimensional variables and nonlinear, we extend our D$^2$VD to a non-linear version, namely Nonlinear-D$^2$VD (N-D$^2$VD) algorithm. To validate the effectiveness of our proposed algorithms, we conduct extensive experiments on both synthetic and real-world datasets. The experimental results demonstrate that our D$^2$VD and N-D$^2$VD algorithms can automatically separate the variables precisely, and estimate treatment effect more accurately and with tighter confidence intervals than the state-of-the-art methods. We also demonstrated that the top-ranked features by our algorithm have the best prediction performance on an online advertising dataset.

Adversarial training is one of the most effective approaches to improve model robustness against adversarial examples. However, previous works mainly focus on the overall robustness of the model, and the in-depth analysis on the role of each class involved in adversarial training is still missing. In this paper, we propose to analyze the class-wise robustness in adversarial training. First, we provide a detailed diagnosis of adversarial training on six benchmark datasets, i.e., MNIST, CIFAR-10, CIFAR-100, SVHN, STL-10 and ImageNet. Surprisingly, we find that there are remarkable robustness discrepancies among classes, leading to unbalance/unfair class-wise robustness in the robust models. Furthermore, we keep investigating the relations between classes and find that the unbalanced class-wise robustness is pretty consistent among different attack and defense methods. Moreover, we observe that the stronger attack methods in adversarial learning achieve performance improvement mainly from a more successful attack on the vulnerable classes (i.e., classes with less robustness). Inspired by these interesting findings, we design a simple but effective attack method based on the traditional PGD attack, named Temperature-PGD attack, which proposes to enlarge the robustness disparity among classes with a temperature factor on the confidence distribution of each image. Experiments demonstrate our method can achieve a higher attack rate than the PGD attack. Furthermore, from the defense perspective, we also make some modifications in the training and inference phase to improve the robustness of the most vulnerable class, so as to mitigate the large difference in class-wise robustness. We believe our work can contribute to a more comprehensive understanding of adversarial training as well as rethinking the class-wise properties in robust models.

Centralized Training with Decentralized Execution (CTDE) has been a popular paradigm in cooperative Multi-Agent Reinforcement Learning (MARL) settings and is widely used in many real applications. One of the major challenges in the training process is credit assignment, which aims to deduce the contributions of each agent according to the global rewards. Existing credit assignment methods focus on either decomposing the joint value function into individual value functions or measuring the impact of local observations and actions on the global value function. These approaches lack a thorough consideration of the complicated interactions among multiple agents, leading to an unsuitable assignment of credit and subsequently mediocre results on MARL. We propose Shapley Counterfactual Credit Assignment, a novel method for explicit credit assignment which accounts for the coalition of agents. Specifically, Shapley Value and its desired properties are leveraged in deep MARL to credit any combinations of agents, which grants us the capability to estimate the individual credit for each agent. Despite this capability, the main technical difficulty lies in the computational complexity of Shapley Value who grows factorially as the number of agents. We instead utilize an approximation method via Monte Carlo sampling, which reduces the sample complexity while maintaining its effectiveness. We evaluate our method on StarCraft II benchmarks across different scenarios. Our method outperforms existing cooperative MARL algorithms significantly and achieves the state-of-the-art, with especially large margins on tasks with more severe difficulties.

Learning user representations based on historical behaviors lies at the core of modern recommender systems. Recent advances in sequential recommenders have convincingly demonstrated high capability in extracting effective user representations from the given behavior sequences. Despite significant progress, we argue that solely modeling the observational behaviors sequences may end up with a brittle and unstable system due to the noisy and sparse nature of user interactions logged. In this paper, we propose to learn accurate and robust user representations, which are required to be less sensitive to (attack on) noisy behaviors and trust more on the indispensable ones, by modeling counterfactual data distribution. Specifically, given an observed behavior sequence, the proposed CauseRec framework identifies dispensable and indispensable concepts at both the fine-grained item level and the abstract interest level. CauseRec conditionally samples user concept sequences from the counterfactual data distributions by replacing dispensable and indispensable concepts within the original concept sequence. With user representations obtained from the synthesized user sequences, CauseRec performs contrastive user representation learning by contrasting the counterfactual with the observational. We conduct extensive experiments on real-world public recommendation benchmarks and justify the effectiveness of CauseRec with multi-aspects model analysis. The results demonstrate that the proposed CauseRec outperforms state-of-the-art sequential recommenders by learning accurate and robust user representations.

In recommender systems, modeling user-item behaviors is essential for user representation learning. Existing sequential recommenders consider the sequential correlations between historically interacted items for capturing users’ historical preferences. However, since users’ preferences are by nature time-evolving and diversified, solely modeling the historical preference (without being aware of the time-evolving trends of preferences) can be inferior for recommending complementary or fresh items and thus hurt the effectiveness of recommender systems. In this paper, we bridge the gap between the past preference and potential future preference by proposing the future-aware diverse trends (FAT) framework. By future-aware, for each inspected user, we construct the future sequences from other similar users, which comprise of behaviors that happen after the last behavior of the inspected user, based on a proposed neighbor behavior extractor. By diverse trends, supposing the future preferences can be diversified, we propose the diverse trends extractor and the time-aware mechanism to represent the possible trends of preferences for a given user with multiple vectors. We leverage both the representations of historical preference and possible future trends to obtain the final recommendation. The quantitative and qualitative results from relatively extensive experiments on real-world datasets demonstrate the proposed framework not only outperforms the state-of-the-art sequential recommendation methods across various metrics, but also makes complementary and fresh recommendations.

The rapid development of modern communication technologies-in particular, (mobile) phone communications-has largely facilitated human social interactions and information exchange. However, the emergence of telemarketing frauds can significantly dissipate individual fortune and social wealth, resulting in a potential slow down or damage to economics. In this work, we propose to spot telemarketing frauds, with an emphasis on unveiling the “precise fraud” phenomenon and the strategies that are used by fraudsters to precisely select targets. To study this problem, we employ a one-month complete dataset of telecommunication metadata in Shanghai with 54 million users and 698 million call logs. Through our study, we find that user's information might have been seriously leaked, and fraudsters have a preference over the target user's age and activity in mobile network. We further propose a novel semi-supervised learning framework to distinguish fraudsters from non-fraudsters. Experimental results on a real-world data show that our approach outperforms several state-of-the-art algorithms in accuracy of detecting fraudsters (e.g., +0.278 in terms of F1 on average). We believe that our study can potentially inform policymaking for government and mobile service providers.

In e-commerce, consumer-generated videos, which in general deliver consumers' individual preferences for the different aspects of certain products, are massive in volume. To recommend these videos to potential consumers more effectively, diverse and catchy video titles are critical. However, consumer-generated videos seldom accompany appropriate titles. To bridge this gap, we integrate comprehensive sources of information, including the content of consumer-generated videos, the narrative comment sentences supplied by consumers, and the product attributes, in an end-to-end modeling framework. Although automatic video titling is very useful and demanding, it is much less addressed than video captioning. The latter focuses on generating sentences that describe videos as a whole while our task requires the product-aware multi-grained video analysis. To tackle this issue, the proposed method consists of two processes, i.e., granular-level interaction modeling and abstraction-level story-line summarization. Specifically, the granular-level interaction modeling first utilizes temporal-spatial landmark cues, descriptive words, and abstractive attributes to builds three individual graphs and recognizes the intra-actions in each graph through Graph Neural Networks (GNN). Then the global-local aggregation module is proposed to model inter-actions across graphs and aggregate heterogeneous graphs into a holistic graph representation. The abstraction-level story-line summarization further considers both frame-level video features and the holistic graph to utilize the interactions between products and backgrounds, and generate the story-line topic of the video. We collect a large-scale dataset accordingly from real-world data in Taobao, a world-leading e-commerce platform, and will make the desensitized version publicly available to nourish further development of the research community. Relatively extensive experiments on various datasets demonstrate the efficacy of the proposed method.

In the digital era, information retrieval, text/knowledge mining, and NLP techniques are playing increasingly vital roles in legal domain. While the open datasets and innovative deep learning methodologies provide critical potentials, in the legal-domain, efforts need to be made to transfer the theoretical/algorithmic models into the real applications to assist users, lawyers, judges and the legal professions to solve the real problems. The objective of this workshop is to aggregate studies/applications of text mining/retrieval and NLP automation in the context of classical/novel legal tasks, which address algorithmic, data and social challenges of legal intelligence. Keynote and invited presentations from industry and academic will be able to fill the gap between ambition and execution in the legal domain.

Treatment effect plays an important role on decision making in many fields, such as social marketing, healthcare, and public policy. The key challenge on estimating treatment effect in the wild observational studies is to handle confounding bias induced by imbalance of the confounder distributions between treated and control units. Traditional methods remove confounding bias by re-weighting units with supposedly accurate propensity score estimation under the unconfoundedness assumption. Controlling high-dimensional variables may make the unconfoundedness assumption more plausible, but poses new challenge on accurate propensity score estimation. One strand of recent literature seeks to directly optimize weights to balance confounder distributions, bypassing propensity score estimation. But existing balancing methods fail to do selection and differentiation among the pool of a large number of potential confounders, leading to possible underperformance in many high-dimensional settings. In this article, we propose a data-driven Differentiated Confounder Balancing (DCB) algorithm to jointly select confounders, differentiate weights of confounders and balance confounder distributions for treatment effect estimation in the wild high-dimensional settings. Besides, under some settings with heavy confounding bias, in order to further reduce the bias and variance of estimated treatment effect, we propose a Regression Adjusted Differentiated Confounder Balancing (RA-DCB) algorithm based on our DCB algorithm by incorporating outcome regression adjustment. The synergistic learning algorithms we proposed are more capable of reducing the confounding bias in many observational studies. To validate the effectiveness of our DCB and RA-DCB algorithms, we conduct extensive experiments on both synthetic and real-world datasets. The experimental results clearly demonstrate that our algorithms outperform the state-of-the-art methods. By incorporating regression adjustment, our RA-DCB algorithm achieves more precise estimation on treatment effect than DCB algorithm, especially under the settings with heavy confounding bias. Moreover, we show that the top features ranked by our algorithm generate accurate prediction of online advertising effect.

Microcredit, very small loans given out without any collaterals, is a new form of financial instrument that serves the segment of population that are typically underserved by traditional financial services. When microcredit takes the form of lending over the internet, it has the advantage of easy online application process and fast funding for borrowers, as well as attractive rate of return for individual lenders. For platforms that facilitate such activities, the key challenge lies in risk management, i.e. adequately pricing each loan's risk so as to balance borrowers' lending cost and lenders' risk-adjusted return. In fact, identifying default borrowers is of critical importance for the ecosystem. Traditionally, credit risk depends heavily on borrowers' historical loan records. However, most borrowers do not have any bureau history, and therefore cannot provide sufficient loan records. In this paper, we study default prediction in online lending by using social behavior. Specifically, we based our work on a dataset provided by PPDai, one of the leading platforms in China. Our dataset consists of over 11 million users and more than 1.5 billion call logs between them. We establish a mobile network and explore social factors that predict borrowers' default. Based on this, we focused on cheating agents, who recruit and teach borrowers to cheat by providing false information and faking application materials. Cheating agents represent a type of default, especially detrimental to the system. We propose a novel probabilistic framework to identify default borrowers and cheating agents simultaneously. Experimental results on production dataset demonstrate significant improvement over several baseline methods. Moreover, our model can effectively identify cheating agents without any labels.

Multi-role court debate is a critical component in a civil trial where parties from different camps (plaintiff, defendant, witness, judge, etc.) actively involved. Unlike other types of dialogue, court debate can be lengthy, and important information, with respect to the controversy focus(es), often hides within the redundant and colloquial dialogue data. Summarizing court debate can be a novel but significant task to assist judge to effectively make the legal decision for the target trial. In this work, we propose an innovative end-to-end model to address this problem. Unlike prior summarization efforts, the proposed model projects the multi-role debate into the controversy focus space, which enables high-quality essential utterance(s) extraction in terms of legal knowledge and judicial factors. An extensive set of experiments with a large civil trial dataset shows that the proposed model can provide more accurate and readable summarization against several alternatives in the multi-role court debate scene.

Different time series is measured in almost all fields including biology, economics and sociology. A common challenge for using such data is the imputation of the missing values with reasonable ones. Most of existing approaches to data imputation assume that individual's observations are independent to each other, which is rarely the case in real-world. In this paper, we study the social-aware time series imputation problem. Given a social network that represents social relations between individuals, we propose a sequential encoder-decoder-based framework and build a connection between the missing observations and the social context. In particular, the proposed model employs the attention mechanism to incorporate social context and temporal context into the imputation task. Experimental results based on two real-world datasets demonstrate that our approach outperforms 11 different baseline methods.

Predicting traffic conditions from online route queries is a challenging task as there are many complicated interactions over the roads and crowds involved. In this paper, we intend to improve traffic prediction by appropriate integration of three kinds of implicit but essential factors encoded in auxiliary information. We do this within an encoder-decoder sequence learning framework that integrates the following data: 1) offline geographical and social attributes. For example, the geographical structure of roads or public social events such as national celebrations; 2) road intersection information. In general, traffic congestion occurs at major junctions; 3) online crowd queries. For example, when many online queries issued for the same destination due to a public performance, the traffic around the destination will potentially become heavier at this location after a while. Qualitative and quantitative experiments on a real-world dataset from Baidu have demonstrated the effectiveness of our framework.

Much work has been done recently on learning word embeddings from large corpora, which attempts to find the coordinates of words in a static and high dimensional semantic space. In reality, such corpora often span a sufficiently long time period, during which the meanings of many words may have changed. The co-evolution of word meanings may also result in a distortion of the semantic space, making these static embeddings unable to accurately represent the dynamics of semantics. In this paper, we present a novel computational method to capture such changes and to model the evolution of word semantics. Distinct from existing approaches that learn word embeddings independently from time periods and then align them, our method explicitly establishes the stable topological structure of word semantics and identifies the surprising changes in the semantic space over time through a principled statistical method. Empirical experiments on large-scale real-world corpora demonstrate the effectiveness of the proposed approach, which outperforms the state-of-the-art by a large margin.

In China, 260 million people migrate to cities to realize their urban dreams. Despite that these migrants play an important role in the rapid urbanization process, many of them fail to settle down and eventually leave the city. The integration process of migrants thus raises an important issue for scholars and policymakers. In this paper, we use Shanghai as an example to investigate migrants' behavior in their first weeks and in particular, how their behavior relates to early departure. Our dataset consists of a one-month complete dataset of 698 telecommunication logs between 54 million users, plus a novel and publicly available housing price data for 18K real estates in Shanghai. We find that migrants who end up leaving early tend to neither develop diverse connections in their first weeks nor move around the city. Their active areas also have higher housing prices than that of staying migrants. We formulate a churn prediction problem to determine whether a migrant is going to leave based on her behavior in the first few days. The prediction performance improves as we include data from more days. Interestingly, when using the same features, the classifier trained from only the first few days is already as good as the classifier trained using full data, suggesting that the performance difference mainly lies in the difference between features.

How to improve the quality of conversations in online communities has attracted considerable attention recently. Having engaged, civil, and reactive online conversations has a critical effect on the social life of Internet users. In this study, we are particularly interested in identifying a post in a multi-party conversation that is unlikely to be further replied to, which therefore kills that thread of the conversation. For this purpose, we propose a deep learning model called the ConverNet. ConverNet is attractive due to its capability of modeling the internal structure of a long conversation and its appropriate encoding of the contextual information of the conversation, through effective integration of attention mechanisms. Empirical experiments on real-world datasets demonstrate the effectiveness of the proposed model. For the widely concerned topic, our analysis also offers implications for how to improve the quality and user experience of online conversations, or how to engage users in a conversation with a chatbot.

Distant Supervision is a widely used approach for training relation extraction models. It generates noisy training samples by heuristically labeling a corpus using an existing knowledge base. Previous noise reduction methods for distant supervision fail to utilize information such as data credibility and sample confidence. In this paper, we proposed a novel neural framework, named ENCORE (External Neural COnstraints REgularized distant supervision), which allows an integration of other information for standard DS through regularizations under multiple external neural networks. In ENCORE, a teacher-student co-training mechanism is used to iterative distilling information from external neural networks to an existing relation extraction model. The experiment results demonstrated that without increasing any data or reshaping its original structure, ENCORE enhanced a CNN based relation extraction model for over 12%. The enhanced model also outperforms the state-of-the-art relation extraction method on the same dataset.

With the rapid development of mobile devices, point-of-interest (POI) suggestion has become a popular online web service, which provides attractive and interesting locations to users. In order to provide interesting POIs, many existing POI recommendation works learn the latent representations of users and POIs from users' past visiting POIs, which suffers from the sparsity problem of POI data. In this paper, we consider the problem of POI suggestion from the viewpoint of learning geosocial multimedia network representations. We propose a novel max-margin metric geosocial multimedia network representation learning framework by exploiting users' check-in behavior and their social relations. We then develop a random-walk based learning method with max-margin metric network embedding. We evaluate the performance of our method on a large-scale geosocial multimedia network dataset and show that our method achieves the best performance than other state-of-the-art solutions.

The query logs from an on-line map query system provide rich cues to understand the behaviors of human crowds. With the growing ability of collecting large scale query logs, the query suggestion has been a topic of recent interest. In general, query suggestion aims at recommending a list of relevant queries w.r.t. users’ inputs via an appropriate learning of crowds’ query logs. In this paper, we are particularly interested in map query suggestions (e.g., the predictions of location-related queries) and propose a novel modelHierarchical Contextual Attention Recurrent Neural Network(HCAR-NN) for map query suggestion in an encoding-decoding manner. Given crowds map query logs, our proposed HCAR-NN not only learns the local temporal correlation among map queries in a query session (e.g., queries in a short-term interval are relevant to accomplish a search mission), but also captures the global longer range contextual dependencies among map query sessions in query logs (e.g., how a sequence of queries within a short-term interval has an influence on another sequence of queries). We evaluate our approach over millions of queries from a commercial search engine (i.e.,Baidu Map). Experimental results show that the proposed approach provides significant performance improvements over the competitive existing methods in terms of classical metrics (i.e.,Recall@KandMRR) as well as the prediction of crowds’ search missions.

During the last decade, community-based question answering (CQA) sites have accumulated a vast amount of questions and their crowdsourced answers over time. How to efficiently identify the quality of answers that are relevant to a given question has become an active line of research in CQA. The major challenge of CQA is the accurate selection of high-quality answers w.r.t given questions. Previous approaches tend to model the semantic matching between individual pair of one question and its corresponding answer (how fitting an answer is to a posted question). However, these works ignore the temporal interactions between answers (how previous answers influence the late posted answers). For example, a rational user likely adapts others' opinions, revises his inclinations, and posts a more appropriate answer after understanding the given question and previously posted answers. As a result, this paper devises an architecture named Temporal Interaction and Causal Influence LSTM (TC-LSTM) to effectively leverage not only the causal influence between question-answer (how appropriate an answer is for a given question) but also the temporal interactions between answers-answer (how a high-quality answer gradually forms). In particular, long short-term memory (LSTM) is used to capture the explicit question-answer influence and the implicit answers-answer interactions. Experiments are conducted on SemEval 2015 CQA dataset for answer classification task and Baidu Zhidao Dataset for answer ranking task. The experimental results show the advantage of our model comparing with other state-of-the-art methods.

Many of the words in a given document either deliver facts (objective) or express opinions (subjective), respectively, depending on the topics they are involved in. For example, given a bunch of documents, the word “bug” assigned to the topic “order Hemiptera” apparently remarks one object (i.e., one kind of insects), while the same word assigned to the topic “software” probably conveys a negative opinion. Motivated by the intuitive assumption that different words have varying degrees of discriminative power in delivering the objective sense or the subjective sense with respect to their assigned topics, a model named as discriminatively objective-subjective LDA (dosLDA) is proposed in this paper. The essential idea underlying the proposed dosLDA is that a pair of objective and subjective selection variables are explicitly employed to encode the interplay between topics and discriminative power for the words in documents in a supervised manner. As a result, each document is appropriately represented as “bag-of-discriminativewords” (BoDW). The experiments reported on documents and images demonstrate that dosLDA not only performs competitively over traditional approaches in terms of topic modeling and document classification, but also has the ability to discern the discriminative power of each word in terms of its objective or subjective sense with respect to its assigned topic.

We propose selective supervised Latent Dirichlet Allocation (ssLDA) to boost the prediction performance of the widely studied supervised probabilistic topic models. We introduce a Bernoulli distribution for each word in one given document to selectthis word as a strongly or weakly discriminative one with respect to its assigned topic. The Bernoulli distribution is parameterized by the discrimination power of the word for its assigned topic. As a result, the document is represented as a “bag-of-selective-words” instead of the probabilistic “bag-of-topics” in the topic modeling domain or the flat “bag-of-words” in the traditional natural language processing domain to form a new perspective. Inheriting the general framework of supervised LDA (sLDA), ssLDA can also predict many types of response specified by a Gaussian Linear Model (GLM). Focusing on the utilization of this word selection mechanism for singe-label document classification in this paper, we conduct the variational inference for approximating the intractable posterior and derive a maximum-likelihood estimation of parameters in ssLDA. The experiments reported on textual documents show that ssLDA not only performs competitively over “state-of-the-art” classification approaches based on both the flat “bag-of-words” and probabilistic “bag-of-topics” representation in terms of classification performance, but also has the ability to discover the discrimination power of the words specified in the topics (compatible with our rational knowledge).

Cross-media hashing, which conducts cross-media retrieval by embedding data from different modalities into a common low-dimensional Hamming space, has attracted intensive attention in recent years. The existing cross-media hashing approaches only aim at learning hash functions to preserve the intra-modality and inter-modality correlations, but do not directly capture the underlying semantic information of the multi-modal data. We propose a discriminative coupled dictionary hashing (DCDH) method in this paper. In DCDH, the coupled dictionary for each modality is learned with side information (e.g., categories). As a result, the coupled dictionaries not only preserve the intra-similarity and inter-correlation among multi-modal data, but also contain dictionary atoms that are semantically discriminative (i.e., the data from the same category is reconstructed by the similar dictionary atoms). To perform fast cross-media retrieval, we learn hash functions which map data from the dictionary space to a low-dimensional Hamming space. Besides, we conjecture that a balanced representation is crucial in cross-media retrieval. We introduce multi-view features on the relatively ``weak'' modalities into DCDH and extend it to multi-view DCDH (MV-DCDH) in order to enhance their representation capability. The experiments on two real-world data sets show that our DCDH and MV-DCDH outperform the state-of-the-art methods significantly on cross-media retrieval.

Hashing techniques have been extensively investigated to boost similarity search for large-scale high-dimensional data. Most of the existing approaches formulate the their objective as a pair-wise similarity-preserving problem. In this paper, we consider the hashing problem from the perspective of optimizing a list-wise learning to rank problem and propose an approach called List-Wise supervised Hashing (LWH). In LWH, the hash functions are optimized by employing structural SVM in order to explicitly minimize the ranking loss of the whole list-wise permutations instead of merely the point-wise or pair-wise supervision. We evaluate the performance of LWH on two real-world data sets. Experimental results demonstrate that our method obtains a significant improvement over the state-of-the-art hashing approaches due to both structural large margin and list-wise ranking pursuing in a supervised manner.

Cross-modal retrieval is a classic research topic in multimedia information retrieval. The traditional approaches study the problem as a pairwise similarity function problem. In this paper, we consider this problem from a new perspective as a listwise ranking problem and propose a general cross-modal ranking algorithm to optimize the listwise ranking loss with a low rank embedding, which we call Latent Semantic Cross-Modal Ranking (LSCMR). The latent low-rank embedding space is discriminatively learned by structural large margin learning to optimize for certain ranking criteria directly. We evaluate LSCMR on the Wikipedia and NUS-WIDE dataset. Experimental results show that this method obtains significant improvements over the state-of-the-art methods.

Recently, there has been extensive research towards the user-provided tags on photo sharing websites which can greatly facilitate image retrieval and management. However, due to the arbitrariness of the tagging activities, these tags are often imprecise and incomplete. As a result, quite a few technologies has been proposed to improve the user experience on these photo sharing systems, including tag clustering and refinement, etc. In this work, we propose a novel framework to model the relationships among tags and images which can be applied to many tag based applications. Different from previous approaches which model images and tags as heterogeneous objects, images and their tags are uniformly viewed as compositions of Semantic Unities in our framework. Then Semantic Unity Graph (SUG) is introduced to represent the complex and high-order relationships among these Semantic Unities. Based on the representation of Semantic Unity Graph, the relevance of images and tags can be naturally measured in terms of the similarity of their Semantic Unities. Then Tag clustering and refinement can then be performed on SUG and the polysemy of images and tags is explicitly considered in this framework. The experiment results conducted on NUS-WIDE and MIR-Flickr datasets demonstrate the effectiveness and efficiency of the proposed approach.