Recommendation is crucial for both user experience and company revenue in Meituan as a leading lifestyle company, and generative recommendation models (GRMs) are shown to produce quality recommendations recently. However, existing systems are limited by insufficient functionality support and inefficient implementations for training GRMs in industrial scenarios. As such, we introduce MTGenRec as an efficient and scalable system for GRM training. Specifically, to handle real-time insertions/deletions of sparse embeddings, MTGenRec employs dynamic hash tables to replace static ones. To improve training efficiency, MTGenRec conducts dynamic sequence balancing to address the computation load imbalances among GPUs and adopts feature ID deduplication alongside automatic table merging to accelerate embedding lookup. Extensive experiments show that MTGenRec improves training throughput by 1.6x - 2.4x while achieving good scalability when running over 100 GPUs. MTGenRec has been deployed for many applications in Meituan and is now handling hundreds of millions of requests on a daily basis. On the delivery platform, we observe a 1.22% growth in user order volume and a 1.31% enhancement in online PV_CTR.

Abstract The rapid proliferation of images on online platforms has made emotion analysis a task of paramount significance. However, these images are often privacy-sensitive, making Federated Learning (FL) a compelling paradigm over traditional centralized methods. A critical yet largely unaddressed challenge in applying FL to this domain is the severe concept drift stemming from the subjective and culturally diverse nature of emotional expression, which causes conventional FL algorithms to fail. In this paper, we propose CAFL (Conditional Attention Federated Learning) to fill this gap. CAFL empowers clients to learn collaboratively yet personally. It intelligently routes information through an adaptive gate that separates features into a personalized stream and a global stream. These streams are then processed by dedicated local and global prediction heads. Crucially, collaboration is guided by a conditional attention mechanism, where the server computes a personalized reference model for each client based on an attention-weighted aggregation of peer models, promoting knowledge sharing among kindred clients. Extensive experiments on various lightweight foundation models show that CAFL consistently outperforms existing FL methods, demonstrating its robustness and superior performance as a solution for distributed, privacy-sensitive image emotion analysis.

Due to the rising concerns on privacy protection, how to build machine learning models from distributed databases with privacy guarantees has gained more popularity. Vertical federated learning (VFL) trains machine learning models in a privacy-preserving way when the data features are scattered over distributed databases. We study the participant selection problem (PSP) for VFL, which chooses a given number of participants to conduct training while maximizing model accuracy. Compared to training with all participants, PSP can filter out hitch-riders that contribute marginally to model quality and reduce training time by involving fewer participants. To achieve good model accuracy, we formulate PSP as choosing a set of participants that maximizes the likelihood of the data samples. Then, utilizing the k-nearest neighbors (KNN) classifier as the proxy model, we express the likelihood as a function of the selected participants and prove that the function is sub modular. The submodular property is favorable as it can account for the feature diversity among the participants and allows to greedily select the participant with the maximum gain in each step. However, the selection process requires finding the top-k neighbors of a data sample as the basic operation, which is expensive in VFL setting as it involves encrypted communication. As such, we adapt the Fagin's algorithm, a famous top-k query algorithm, to reduce the amount of encrypted communication. We deploy our solution VFPS-SM across five distributed nodes and conduct experiments with 10 datasets and 3 models to evaluate its performance. The results show that VFPS-SM can reduce the end-to-end running time by up to$35\times$, selection time$365\times$and improve model accuracy by 6.0% compared with state-of-the-art baselines.

Mobile phone agents can assist people in automating daily tasks on their phones, which have emerged as a pivotal research spotlight. However, existing procedure-oriented agents struggle with cross-app instructions, due to the following challenges: (1) complex task relationships, (2) diverse app environment, and (3) error propagation and information loss in multi-step execution. Drawing inspiration from object-oriented programming principles, we recognize that object-oriented solutions is more suitable for cross-app instruction. To address these challenges, we propose a self-evolving multi-agent framework named MobileSteward which integrates multiple app-oriented StaffAgents coordinated by a centralized StewardAgent. We design three specialized modules in MobileSteward: (1) Dynamic Recruitment generates a scheduling graph guided by information flow to explicitly associate tasks among apps. (2) Assigned Execution assigns the task to app-oriented StaffAgents, each equipped with app-specialized expertise to address the diversity between apps. (3) Adjusted Evaluation conducts evaluation to provide reflection tips or deliver key information, which alleviates error propagation and information loss during multi-step execution. To continuously improve the performance of MobileSteward, we develop a Memory-based Self-evolution mechanism, which summarizes the experience from successful execution, to improve the performance of MobileSteward. We establish the first English Cross-APP Benchmark (CAPBench) in the real-world environment to evaluate the agents' capabilities of solving complex cross-app instructions. Experimental results demonstrate that MobileSteward achieves the best performance compared to both single-agent and multi-agent frameworks, highlighting the superiority of MobileSteward in better handling user instructions with diverse complexity.

Due to the lack of an effective auditing mechanism for malicious participants, federated learning (FL) framework faces serious threats of Byzantine attacks. Existing Byzantine resilience methods generally use the same batch size by default. However, this same batch size setting may not be applicable in practice, since clients may comprise a diverse array of devices with different data storage and computational capabilities. This paper mainly studies the robustness of Byzantine across different clients with different batch sizes in FL systems. Specifically, we propose a weighted aggregation framework (WAggF) to enhance the Byzantine robustness of the FL systems in case of different clients with different batch sizes. Combining WAggF with Byzantine attack resilient distributed (Byrd-) algorithms, we develop two Byzantine-robustness algorithms: Byrd-SGD with Different Batch sizes (Byrd-DBSGD) and Byrd-SAGA with Different Batch sizes (Byrd-DBSAGA). Theoretically, we prove that these algorithms exhibit superior performance compared to their original counterparts in the presence of Byzantine attacks. Additionally, we introduce a data partition method to solve the over-centralization problem caused by significantly disparate different batch sizes among different clients. Considering the communication efficiency, we utilize an unbiased compressor to ensure synchronous training of the FL system. To the best of our knowledge, this study represents the first investigation into the impact of different batch sizes across different clients on Byzantine robustness in FL. Extensive experimental results confirm that our proposed algorithms exhibit greater resilience to Byzantine attacks. Codes will be released upon publication.

Abstract Link prediction is essential for identifying hidden relationships within network data, with significant implications for fields such as social network analysis and bioinformatics. Traditional methods often overlook potential relationships among common neighbors, limiting their effectiveness in utilizing graph information fully. To address this, we introduce a novel approach, Common Neighbor Completion with Information Entropy (IECNC), which enhances model expressiveness by considering logical neighbor relationships. Our method integrates a dynamic node function with a Message Passing Neural Network (MPNN), focusing on first-order neighbors and employing set-based aggregation to improve missing link predictions. By combining the information entropy of probabilistic predictions of common neighbors with MPNN and leveraging information entropy to assess uncertainty in adjacent connections, our approach significantly enhances prediction accuracy. Experimental results demonstrate that our IECNC method achieves optimal performance across multiple datasets, surpassing existing techniques. Furthermore, visualizations confirm that our model effectively captures and accurately learns feature information from various categories, Demonstrating the method’s efficacy and adaptability.

Multi-modal contents have proven to be the powerful knowledge for recommendation tasks. Most state-of-the-art multi-modal recommendation methods mainly focus on aligning the semantic spaces of different modalities to enhance the item representations and do not pay much attention on the relevant knowledge in the multi-modalities for recommendation, resulting in that the positive effects of the relevant knowledge is reduced and the improvement of recommendation performance is limited. In this paper, we propose a multi-modal correction network termed MMCN to enhance the item representation with the important semantic knowledge in each modality by a residual structure with attention mechanisms and a hierarchical contrastive learning framework. The residual information is obtained through self-attention and cross-attention, which can learn the relevant knowledge across different modalities effectively. While hierarchical contrastive learning further captures the relevant knowledge not only at the feature level but also at the element-wise level with a matrix. Extensive experiments on three large-scale real-world datasets show the superiority of MMCN over state-of-the-art multi-modal recommendation methods.

For graph self-supervised learning (GSSL), masked autoencoder (MAE) follows the generative paradigm and learns to reconstruct masked graph edges or node features while contrastive learning (CL) maximizes the similarity between augmented views of the same graph. Existing works utilize MAE and CL separately but we observe that the MAE and CL paradigms are complementary and propose the graph contrastive masked autoencoder (GCMAE) framework to unify them. Specifically, by focusing on local edges or node features, MAE cannot capture global information of the graph and is sensitive to particular edges and features. On the contrary, CL excels in extracting global information because it considers the relation between graphs. As such, we equip GCMAE with an MAE branch and a CL branch, and the two branches share a common encoder, which allows the MAE branch to exploit the global information extracted by the CL branch. To force GCMAE to capture global graph structures, we train it to reconstruct the entire adjacency matrix instead of only the masked edges as in existing works. Moreover, a discrimination loss is proposed for feature reconstruction, which improves the disparity between node embeddings rather than reducing the reconstruction error to tackle the feature smoothing problem of MAE. We evaluate GCMAE on four popular graph tasks (i.e., node classification, node clustering, link prediction, and graph classification) and compare it with 14 state-of-the-art baselines. The results show that GCMAE consistently provides good accuracy across these tasks, and the maximum accuracy improvement is up to 3.2% compared with the best-performing baseline.

Graph dataset condensation (GDC) reduces a dataset with many graphs into a smaller dataset with fewer graphs while maintaining model training accuracy. GDC saves the storage cost and hence accelerates training. Although several GDC methods have been proposed, they are all supervised and require massive labels for the graphs, while graph labels can be scarce in many practical scenarios. To fill this gap, we propose a self-supervised graph dataset condensation method called SGDC, which does not require label information. Our initial design starts with the classical bilevel optimization paradigm for dataset condensation and incorporates contrastive learning techniques. But such a solution yields poor accuracy due to the biased gradient estimation caused by data augmentation. To solve this problem, we introduce representation matching, which conducts training by aligning the representations produced by the condensed graphs with the target representations generated by a pre-trained SSL model. This design eliminates the need for data augmentation and avoids biased gradient. We further propose a graph attention kernel, which not only improves accuracy but also reduces running time when combined with self-supervised kernel ridge regression (KRR). To simplify SGDC and make it more robust, we adopt a adjacency matrix reusing approach, which reuses the topology of the original graphs for the condensed graphs instead of repeatedly learning topology during training. Our evaluations on seven graph datasets find that SGDC improves model accuracy by up to 9.7% compared with 5 state-of-the-art baselines, even if they use label information. Moreover, SGDC is significantly more efficient than the baselines.

In recent years, uplift modeling, also known as individual treatment effect (ITE) estimation, has seen wide applications in online marketing, such as delivering one-time issuance of coupons or discounts to motivate users' purchases. However, complex yet more realistic scenarios involving multiple interventions over time on users are still rarely explored. The challenges include handling the bias from time-varying confounders, determining optimal treatment timing, and selecting among numerous treatments. In this paper, to tackle the aforementioned challenges, we present a temporal point process-based uplift model (TPPUM) that utilizes users' temporal event sequences to estimate treatment effects via counterfactual analysis and temporal point processes. In this model, marketing actions are considered as treatments, user purchases as outcome events, and how treatments alter the future conditional intensity function of generating outcome events as the uplift. Empirical evaluations demonstrate that our method outperforms existing baselines on both real-world and synthetic datasets. In the online experiment conducted in a discounted bundle recommendation scenario involving an average of 3 to 4 interventions per day and hundreds of treatment candidates, we demonstrate how our model outperforms current state-of-the-art methods in selecting the appropriate treatment and timing of treatment, resulting in a 3.6% increase in application-level revenue.

CTR (Click-Through Rate) prediction plays an essential role in online advertising systems. Most existing works attempt to capture users’ interests from sessions by assuming that behaviors within a session are homogeneous. However, user interest may change frequently. Thus it is hard to guarantee that behaviors in a session are homogeneous, resulting in users’ interests extracted from sessions being biased. In this paper, we propose a model named Deep Session Heterogeneity-aware Network (DSHN) by learning the relationships of behaviors within sessions and the relevance between the session and target item to alleviate the influence of irrelevant or heterogeneous sessions. We design a heterogeneity-aware mechanism to learn the heterogeneity of items within a session. Then we further design two modules: the Session Heterogeneity Learning module and the Relevance Inference module. The Session Heterogeneity Learning module weighs each session by summarizing the variation of session interest with and without any behavior. The relevance Inference module learns the relevance between the target item and each session in a similar way by learning session interest with and without the target item. Extensive experiments on four datasets demonstrate that our proposed DSHN achieves better results compared to the state-of-the-art.

Prompt Transfer (PoT) is a recently-proposed approach to improve prompt-tuning, by initializing the target prompt with the existing prompt trained on similar source tasks. However, such a vanilla PoT approach usually achieves sub-optimal performance, as (i) the PoT is sensitive to the similarity of source-target pair and (ii) directly fine-tuning the prompt initialized with source prompt on target task might lead to forgetting of the useful general knowledge learned from source task. To tackle these issues, we propose a new metric to accurately predict the prompt transferability (regarding (i)), and a novel PoT approach (namelyPanDa) that leverages the knowledge distillation technique to alleviate the knowledge forgetting effectively (regarding (ii)). Extensive and systematic experiments on 189 combinations of 21 source and 9 target datasets across 5 scales of PLMs demonstrate that: 1)our proposed metric works well to predict the prompt transferability; 2)ourPanDaconsistently outperforms the vanilla PoT approach by 2.3% average score (up to 24.1%) among all tasks and model sizes; 3)with ourPanDaapproach, prompt-tuning can achieve competitive and even better performance than model-tuning in various PLM scales scenarios.

Sequence-to-sequence (seq2seq) learning is a popular fashion for large-scale pretraining language models. However, the previous seq2seq pretraining models generally focus on reconstructive objectives on the decoder side and neglect the effect of encoder-side supervision, which we argue may lead to sub-optimal performance. To verify our hypothesis, we first empirically study the functionalities of the encoder and decoder in seq2seq pretrained language models, and find that the encoder takes an important but under-exploitation role than the decoder regarding the downstream performance and neuron activation. Therefore, we propose an encoding-enhanced seq2seq pretraining strategy, namelyE2S2, which improves the seq2seq models via integrating more efficient self-supervised information into the encoders. Specifically, E2S2 adopts two self-supervised objectives on the encoder side from two aspects: 1) locally denoising the corrupted sentence (denoising objective); and 2) globally learning better sentence representations (contrastive objective). With the help of both objectives, the encoder can effectively distinguish the noise tokens and capture high-level (i.e., syntactic and semantic) knowledge, thus strengthening the ability of seq2seq model to accurately achieve the conditional generation. On a large diversity of downstream natural language understanding and generation tasks, E2S2 dominantly improves the performance of its powerful backbone models, e.g., BART and T5. For example, upon BART backbone, we achieve +1.1% averaged gain on the general language understanding evaluation (GLUE) benchmark and +1.75%$F_{0.5}$score improvement on CoNLL2014 dataset. We also provide in-depth analyses to show the improvement stems from better linguistic representation. We hope that our work will foster future self-supervision research on seq2seq language model pretraining.

Recently, the field of generative models has seen a significant advancement with the introduction of Diffusion Probabilistic Models (DPMs). The Denoising Diffusion Implicit Model (DDIM) was designed to reduce computational time by skipping a number of steps in the inference process of DPMs. However, the hand-crafted sampling schedule in DDIM, which relies on human expertise, has its limitations in considering all relevant factors in the sampling process. Additionally, the assumption that all instances should have the same schedule is not always valid. To address these problems, this paper proposes a method that leverages reinforcement learning to automatically search for an optimal sampling schedule for DPMs. This is achieved by a policy network that predicts the next step to visit based on the current state of the noisy image. The optimization of the policy network is accomplished using an episodic actor-critic framework, which incorporates reinforcement learning. Empirical results demonstrate the superiority of our approach over various datasets with different timesteps. We also observe that the trained sampling schedule has a strong generalization ability across different DPM baselines.

Data augmentation has undoubtedly enabled a significant leap forward in training a high-accuracy deep network. Besides the commonly used augmentation to target data, e.g., random cropping, flipping, and rotation, recent works have been dedicated to mining generalized knowledge by using multiple sources. However, along with plentiful data comes the huge data distribution gap between the target and different sources (hybrid shift). To mitigate this problem, existing methods tend to manually annotate more data. Unlike previous methods, this paper focuses on the study of learning deep models by gathering knowledge from multiple sources in a labor-free fashion and further proposes the "Multi-Alignment and Pseudo-Learning'' method, dubbed MAPLE. MAPLE constructs the multi-alignment module, which consists of multiple discriminators to align different data distributions via an adversarial process. In addition, a novel semi-supervised learning (SSL) manner is introduced to further facilitate the utility of our MAPLE. Extensive evaluations conducted on four benchmarks show the effectiveness of the proposed MAPLE, which achieves state-of-the-art performance outperforming existing methods by an obvious margin.

Node classification is a fundamental graph-based task that aims to predict the classes of unlabeled nodes, for which Graph Neural Networks (GNNs) are the state-of-the-art methods. Current GNNs assume that nodes in the training set contribute equally during training. However, the quality of training nodes varies greatly, and the performance of GNNs could be harmed by two types of low-quality training nodes: (1) inter-class nodes situated near class boundaries that lack the typical characteristics of their corresponding classes. Because GNNs are data-driven approaches, training on these nodes could degrade the accuracy. (2) mislabeled nodes. In real-world graphs, nodes are often mislabeled, which can significantly degrade the robustness of GNNs. To mitigate the detrimental effect of the low-quality training nodes, we present CLNode, which employs a selective training strategy to train GNN based on the quality of nodes. Specifically, we first design a multi-perspective difficulty measurer to accurately measure the quality of training nodes. Then, based on the measured qualities, we employ a training scheduler that selects appropriate training nodes to train GNN in each epoch. To evaluate the effectiveness of CLNode, we conduct extensive experiments by incorporating it in six representative backbone GNNs. Experimental results on real-world networks demonstrate that CLNode is a general framework that can be combined with various GNNs to improve their accuracy and robustness.

Link prediction has attracted considerable attention. Empiricism and the evolution mechanism based approach are the mainstream methods for link prediction. However, one drawback of such approaches is that they usually ignore the dynamic evolution mechanism of social networks, yet being dynamic is an essential characteristic of a social network that exists in every stage of the networks evolution. In this paper, we address the problem of temporal link prediction and investigate social networks from the time dimension with the purpose of dynamic evolution mechanism capturing. First, we separate a temporal network into a series of snapshots. Then, we propose a triad transition matrix prediction algorithm to learn the change of the distribution of triads among the different snapshots. The learned changes in the distribution of triads can capture the dynamic evolution of the network. With a proposed triad transition influence quantification algorithm, we propose a motifs based link prediction method for temporal link prediction. The proposed method can capture the dynamic evolution of temporal networks and is universal than existing methods. Extensive experiments on disparate real-world networks and model networks with controllable evolution demonstrate the effectiveness of the proposed method.

Aspect-based sentiment analysis (ABSA) is a fine-grained task of sentiment analysis. To better comprehend long complicated sentences and obtain accurate aspect-specific information, linguistic and commonsense knowledge are generally required in this task. However, most current methods employ complicated and inefficient approaches to incorporate external knowledge, e.g., directly searching the graph nodes. Additionally, the complementarity between external knowledge and linguistic information has not been thoroughly studied. To this end, we propose a knowledge graph augmented network (KGAN), which aims to effectively incorporate external knowledge with explicitly syntactic and contextual information. In particular, KGAN captures the sentiment feature representations from multiple different perspectives,i.e., context-, syntax- and knowledge-based. First, KGAN learns the contextual and syntactic representations in parallel to fully extract the semantic features. Then, KGAN integrates the knowledge graphs into the embedding space, based on which the aspect-specific knowledge representations are further obtained via an attention mechanism. Last, we propose a hierarchical fusion module to complement these multi-view representations in alocal-to-globalmanner. Extensive experiments on five popular ABSA benchmarks demonstrate the effectiveness and robustness of our KGAN. Notably, with the help of the pretrained model of RoBERTa, KGAN achieves a new record of state-of-the-art performance among all datasets.

Signed network embedding (SNE) has received considerable attention in recent years. A mainstream idea of SNE is to learn node representations by estimating the ratio of sampling densities. Though achieving promising performance, these methods based on density ratio estimation are limited to the issues of confusing sample, expected error, and fixed priori. To alleviate the above-mentioned issues, in this paper, we propose a novel dual-branch density ratio estimation (DDRE) architecture for SNE. Specifically, DDRE 1) consists of a dual-branch network, dealing with the confusing sample; 2) proposes the expected matrix factorization without sampling to avoid the expected error; and 3) devises an adaptive cross noise sampling to alleviate the fixed priori. We perform sign prediction and node classification experiments on four real-world and three artificial datasets, respectively. Extensive empirical results demonstrate that DDRE not only significantly outperforms the methods based on density ratio estimation but also achieves competitive performance compared with other types of methods such as graph likelihood, generative adversarial networks, and graph convolutional networks. Code is publicly available at https://github.com/WHU-SNA/DDRE.

Many studies focusing on integrating reviews with ratings to improve recommendation performance have been quite successful. However, these works still face several shortcomings: (1) The importance of dynamically integrating review and interaction data features is typically ignored, yet treating these fusion features equally may lead to an incomplete understanding of user preferences. (2) Some forms of soft attention methods are adopted to model the local semantic information of words. As features thus captured may contain irrelevant information, the generated attention map is neither discriminatory nor detailed. In this paper, we propose a novelAdaptiveHierarchicalAttention-enhancedGated network integrating reviews for item recommendation, named AHAG. AHAG is a unified framework to capture the hidden intentions of users by adaptively incorporating reviews. Specifically, we design a gated network to dynamically fuse the extracted features and select the features that are most relevant to user preferences. To capture distinguishing fine-grained features, we introduce a hierarchical attention mechanism to learn important semantic information features and the dynamic interaction of these features. Besides, the high-order non-linear interaction of neural factorization machines is utilized to derive the rating prediction. Experiments on seven real-world datasets show that the proposed AHAG significantly outperforms state-of-the-art methods. Furthermore, the attention mechanism can highlight the relevant information in reviews to increase the interpretability of the recommendation task. Source codes are available inhttps://github.com/luojia527/AHAG.

Recent researches indicate that the impact of adversarial perturbations on deep learning models is reflected not only on the alteration of predicted labels but also on the distortion of data structure in the representation space. Significant improvement of the model’s adversarial robustness can be achieved by reforming the structure-aware representation distortion. Current methods generally utilize the one-to-one representation alignment or the triplet information between the positive and negative pairs. However, in this paper, we show that the representation structure of the natural and adversarial examples cannot be well and stably captured if we only focus on a localized range of contrastive examples. To achieve better and more stable adversarial robustness, we propose to adjust the adversarial distortion of representation structure by using Massive Contrastive Adversaries (MCA). Inspired by the Noise-Contrastive Estimation (NCE), MCA exploits the contrastive information by employing m negative instances. Compared with existing methods, our method recruits a much wider range of negative examples per update, so a better and more stable representation relationship between the natural and adversarial examples can be captured. Theoretical analysis shows that the proposed MCA inherently maximizes a lower bound of the mutual information (MI) between the representations of the natural and adversarial examples. Empirical experiments on benchmark datasets demonstrate that MCA can achieve better and more stable intra-class compactness and inter-class divergence, which further induces better adversarial robustness.

Due to the lack of the labeled data and the complex structures of various data, it is very hard to learn the uncertainty and representativeness accurately in active learning. In this paper, we propose a multiple kernel active learning framework that incorporates a group regularizer of distribution information into the estimation of uncertainty. The proposed method takes the advantage of multiple kernel learning to learn the kernel space in which the complex structures can be well captured by kernel weights. Meanwhile, we have developed an efficient optimization algorithm to solve the proposed method. Experimental results on twelve UCI benchmark data sets and eight subsets of ImageNet show that the proposed method outperforms several state-of-the-art active learning methods. Moreover, we also have applied the proposed method to multiple feature scenario on Caltech101, and the promising results are also obtained compared with single feature scenario.

RISC-V is an open-source instruction set and now has been examined as a universal standard to unify the heterogeneous platforms. However, current research focuses primarily on the design and fabrication of general-purpose processors based on RISC-V, despite the fact that in the era of IoT (Internet of Things), the fusion of heterogeneous platforms should also take application-specific processors into account. Accordingly, this paper proposes a collaborative RISC-V multi-core system for Deep Neural Network (DNN) accelerators. To the best of our knowledge, this is the first time that a multi-core scheduling architecture for DNN acceleration is formulated and RISC-V is explored as the ISA of a multi-core system to bridge the gap between the memory and the DNN Processor in order to increase the entire system throughput. The experiment realizes a four-stage design of the RISC-V core, and further reveals that a multi-core design along with an appropriate scheduling algorithm can efficiently decrease the runtime and elevate the throughput. Moreover, the experiment also provides us with a constructive suggestion regarding the ideal proportion of the cores to Process Engines (PE), which provides us with significant assistance in building highly efficient AI System-on-Chips (SoCs) in resource-aware situations.

Graph neural nets are emerging tools to represent network nodes for classification. However, existing approaches typically suffer from two limitations: (1) they only aggregate information from short distance (e.g., 1-hop neighbors) each round and fail to capturelong distance relationship in graphs; (2) they require users to label data from several classes to facilitate the learning of discriminative models; whereas in reality, users may only provide labels of a small number of nodes in a single class. To overcome these limitations, this paper presents a novel long-short distance aggregation networks (\textttLSDAN ) for positive unlabeled (PU) graph learning. Our theme is to generate multiple graphs at different distances based on the adjacency matrix, and further develop a long-short distance attention model for these graphs. The short-distance attention mechanism is used to capture the importance of neighbor nodes to a target node. The long-distance attention mechanism is used to capture the propagation of information within a localized area of each node and help model weights of different graphs for node representation learning. A non-negative risk estimator is further employed, to aggregate long- short-distance networks, for PU learning using back-propagated loss modeling. Experiments on real-world datasets validate the effectiveness of our approach.

Recommender system plays an important role to provide people with personalized information based on their history records. However, it is still a challenge to capture the preference of users accurately due to the sparsity of rating data and the heterogeneity of review data. In this paper, we propose a hybrid deep collaborative filtering model that jointly learns latent representations from ratings and reviews. Specifically, the model learns the rating feature and textual feature based on ratings and reviews simultaneously. Two embedding layers are employed to learn rating feature for users and items based on the user and item interactions, and two attention-based GRU networks learn context-aware representation from user and item reviews. Then a gating mechanism is used to leverage contributions from rating feature and textual feature. Experimental results on six real-world datasets demonstrate the superior performance of the proposed method over several state-of-the-art methods. Moreover, the keywords in reviews can be highlighted to interpret the predictions with the attention mechanism.

Developing effective network embedding methods for social trust networks (STNs) is a non-trivial problem because two key pieces of information need to be preserved simultaneously: a user's relations to latent factors and the trust transfer patterns that govern what type of relationship will form. In this study, we propose a novel social trust network embedding method (STNE) to address these issues. Specifically, we present a modified Skip-Gram model with negative sampling to jointly learn latent factor features, along with the trust transfer pattern features. Moreover, we define a flexible notion about a user's latent relationships with other users, which generates reliable negative samples for optimization. Extensive experiments on several real-world networks demonstrate the efficacy of the proposed STNE.

Despite the great success of many matrix factorization based collaborative filtering approaches, there is still much space for improvement in recommender system field. One main obstacle is the cold-start and data sparseness problem, requiring better solutions. Recent studies have attempted to integrate review information into rating prediction. However, there are two main problems: (1) most of existing works utilize a static and independent method to extract the latent feature representation of user and item reviews ignoring the correlation between the latent features, which may fail to capture the preference of users comprehensively. (2) there is no effective framework that unifies ratings and reviews. Therefore, we propose a novel d ual a ttention m utual l earning between ratings and reviews for item recommendation, named DAML. Specifically, we utilize local and mutual attention of the convolutional neural network to jointly learn the features of reviews to enhance the interpretability of the proposed DAML model. Then the rating features and review features are integrated into a unified neural network model, and the higher-order nonlinear interaction of features are realized by the neural factorization machines to complete the final rating prediction. Experiments on the five real-world datasets show that DAML achieves significantly better rating prediction accuracy compared to the state-of-the-art methods. Furthermore, the attention mechanism can highlight the relevant information in reviews to increase the interpretability of rating prediction.

Link prediction in signed social networks is an important and challenging problem in social network analysis. To produce the most accurate prediction results, two questions must be answered: (1) Which unconnected node pairs are likely to be connected by a link in future? (2) What will the signs of the new links be? These questions are challenging, and current research seldom well solves both issues simultaneously. Additionally, neutral social relationships, which are common in many social networks can affect the accuracy of link prediction. Yet neutral links are not considered in most existing methods. Hence, in this paper, we propose a s igned l atent f actor (SLF) model that answers both these questions and, additionally, considers four types of relationships: positive, negative, neutral and no relationship at all. The model links social relationships of different types to the comprehensive, but opposite, effects of positive and negative SLFs. The SLF vectors for each node are learned by minimizing a negative log-likelihood objective function. Experiments on four real-world signed social networks support the efficacy of the proposed model.

Investigating correlation between example features and example labels is essential to solve classification problems. However, identification and calculation of the correlation between features and labels can be rather difficult for high-dimensional multi-label data. Both feature embedding and label embedding have been developed to tackle this challenge, and a shared subspace for both labels and features are usually learned by existing embedding methods to simultaneously reduce dimensionality of features and labels. In contrast, this paper suggests to learn separated subspaces for features and labels by maximizing the independence between components in each subspace and maximizing the correlation between these two subspaces. The learned independent label components indicates fundamental combinations of labels in multi-label datasets, which thus helps to reveals the correlation between labels. On the other hand, the learned independent feature components lead to a compact representation of example features. The connections between the proposed algorithm and existing embedding methods have been discussed. Experimental results on real-world multi-label datasets demonstrate the necessity of exploring independence components from multi-label data and the effectiveness of the proposed algorithm.

Recently, social networks have witnessed a massive surge in popularity. A key issue in social network research is network evolution analysis, which assumes that all the autonomous nodes in a social network follow uniform evolution mechanisms. However, different nodes in a social network should have different evolution mechanisms to generate different edges. This is proposed as the underlying idea to ensure the nodes' evolution diversity in this paper. Our approach involves identifying the micro-level node evolution that generates different edges by introducing the existing link prediction methods from the perspectives of nodes. We also propose the edge generation coefficient to evaluate the extent to which an edge's generation can be explained by a link prediction method. To quantify the nodes' evolution diversity, we define the diverse evolution distance. Furthermore, a diverse node adaption algorithm is proposed to indirectly analyze the evolution of the entire network based on the nodes' evolution diversity. Extensive experiments on disparate real-world networks demonstrate that the introduction of the nodes' evolution diversity is important and beneficial for analyzing the network evolution. The diverse node adaption algorithm outperforms other state-of-the-art link prediction algorithms in terms of both accuracy and universality. The greater the nodes' evolution diversity, the more obvious its advantages.

In recent years, various data clustering algorithms have been proposed in the data mining and engineering communities. However, there are still drawbacks in traditional clustering methods which are worth to be further investigated, such as clustering for the high dimensional data, learning an ideal affinity matrix which optimally reveals the global data structure, discovering the intrinsic geometrical and discriminative properties of the data space, and reducing the noises influence brings by the complex data input. In this paper, we propose a novel clustering algorithm called robust dual clustering with adaptive manifold regularization (RDC), which simultaneously performs dual matrix factorization tasks with the target of an identical cluster indicator in both of the original and projected feature spaces, respectively. Among which, the$l_{2,1}$-norm is used instead of the conventional$l_{2}$-norm to measure the loss, which helps to improve the model robustness by relieving the influences by the noises and outliers. In order to better consider the intrinsic geometrical and discriminative data structure, we incorporate the manifold regularization term on the cluster indicator by using a particularly learned affinity matrix which is more suitable for the clustering task. Moreover, a novel augmented lagrangian method (ALM) based procedure is designed to effectively and efficiently seek the optimal solution of the proposed RDC optimization. Numerous experiments on the representative data sets demonstrate the superior performance of the proposed method compares to the existing clustering algorithms.

Deep convolutional networks have achieved successful performance in data mining field. However, training large networks still remains a challenge, as the training data may be insufficient and the model can easily get overfitted. Hence the training process is usually combined with a model regularization. Typical regularizers include weight decay, Dropout, etc. In this paper, we propose a novel regularizer, named Structured Decorrelation Constraint (SDC), which is applied to the activations of the hidden layers to prevent overfitting and achieve better generalization. SDC impels the network to learn structured representations by grouping the hidden units and encouraging the units within the same group to have strong connections during the training procedure. Meanwhile, it forces the units in different groups to learn non-redundant representations by minimizing the cross-covariance between them. Compared with Dropout, SDC reduces the co-adaptions between the hidden units in an explicit way. Besides, we propose a novel approach called Reg-Conv that can help SDC to regularize the complex convolutional layers. Experiments on extensive datasets show that SDC significantly reduces overfitting and yields very meaningful improvements on classification performance (on CIFAR-10 6.22% accuracy promotion and on CIFAR-100 9.63% promotion).

The human visual system proves smart in extracting both global and local features. Can we design a similar way for unsupervised feature learning? In this paper, we propose anovel pooling method within an unsupervised feature learningframework, named Rich and Robust Feature Pooling (R2FP), to better explore rich and robust representation from sparsefeature maps of the input data. Both local and global poolingstrategies are further considered to instantiate such a methodand intensively studied. The former selects the most conductivefeatures in the sub-region and summarizes the joint distributionof the selected features, while the latter is utilized to extractmultiple resolutions of features and fuse the features witha feature balancing kernel for rich representation. Extensiveexperiments on several image recognition tasks demonstratethe superiority of the proposed techniques.

In data mining and pattern recognition area, the learned objects are often represented by the multiple features from various of views. How to learn an efficient and effective feature embedding for the subsequent learning tasks? In this paper, we address this issue by providing a novel multi-task multiview feature embedding (MMFE) framework. The MMFE algorithm is based on the idea of low-rank approximation, which suggests that the observed multiview feature matrix is approximately represented by the low-dimensional feature embedding multiplied by a projection matrix. In order to fully consider the particular role of each view to the multiview feature embedding, we simultaneously suggest the multitask learning scheme and ensemble manifold regularization into the MMFE algorithm to seek the optimal projection. Since the objection function of MMFE is multi-variable and non-convex, we further provide an iterative optimization procedure to find the available solution. Two real world experiments show that the proposed method outperforms single-task-based as well as state-of-the-art multiview feature embedding methods for the classification problem.