首页期刊导航|计算机科学技术学报（英文版）
期刊信息/Journal information
计算机科学技术学报（英文版）
计算机科学技术学报（英文版）

李国杰

双月刊

1000-9000

jcst@ict.ac.cn

010-62610746

100080

北京中关村科学院南路6号 《计算机科学技术学报（英）》编辑部

计算机科学技术学报（英文版）/Journal Journal of Computer Science and TechnologyCSCDCSTPCD北大核心EISCI
查看更多>>Journal of Computer Science and Technology(JCST)是中国计算机科学技术领域国际性学术期刊。 JCST于1986 年创刊, 双月刊, 国内外公开发行, 由Springer Science + Business Media代理国际出版发行。 JCST是中国计算机学会会刊, 由中国科学院计算技术研究所承办。JCST由数十位国际计算机界的著名专家和学者联袂编审，把握世界计算机科学技术最新发展趋势。JCST荟萃了国内外计算机科学技术领域中有指导性和开拓性的学术论著，定期组织热点专辑或专题栏目，部分文章邀请了世界著名计算机科学专家撰写。
正式出版
收录年代

    openGauss:An Open-Source Database for the Era of Artificial Intelligence

    李建中
    1005-1006页
      原文链接:
    • 万方数据
    • 维普

    openGauss:An Enterprise-Grade Open-Source Database System

    李国良王江陈国
    1007-1028页
    查看更多>>摘要：We have built openGauss,an enterprise-grade open-source database system.openGauss has fulfilled its de-sign goal of high performance,high availability,high security,and high intelligence.For high performance,it leverages NUMA(non-uniform memory access)-aware data access among multiple cores to enable efficient concurrent transaction processing,and symmetric multi-processing to make use of parallel processing resources adaptively.Moreover,memory-op-timized tables(MOTs)are designed to put everything in memory.For high availability,a three-tier pooling architecture that shares storage among the master and standby instances is proposed to achieve availability at 99.99％,containing both a distributed memory service(DMS)and a distributed storage service(DSS).For high security,it is a fully encrypted database with safe storage features,efficient complex querying,and tamper-proof.For high intelligence,an AI-based opti-mizer in the kernel and a self-driving platform named DBMind are demonstrated to achieve better performance and greater user-friendliness.openGauss has served over 150 enterprises and institutions since its release in 2020.We share the lessons we learned from its development and operation,and our customers.
      原文链接:
    • 万方数据
    • 维普

    Balancing Accuracy and Training Time in Federated Learning for Violence Detection in Surveillance Videos:A Study of Neural Network Architectures

    Quentin PajonSwan SerreHugo WissocqLéo Rabaud...
    1029-1039页
    查看更多>>摘要：This paper presents an original investigation into the domain of violence detection in videos,introducing an innovative approach tailored to the unique challenges of a federated learning environment.The study encompasses a com-prehensive exploration of machine learning techniques,leveraging spatio-temporal features extracted from benchmark video datasets.In a notable departure from conventional methodologies,we introduce a novel architecture,the"Diff Gat-ed"network,designed to streamline preprocessing and training while simultaneously enhancing accuracy.Our exploration of advanced machine learning techniques,such as super-convergence and transfer learning,expands the horizons of federat-ed learning,offering a broader range of practical applications.Moreover,our research introduces a method for seamlessly adapting centralized datasets to the federated learning context,bridging the gap between traditional machine learning and federated learning approaches.The outcome of this study is a remarkable advancement in the field of violence detection,with our federated learning model consistently outperforming state-of-the-art models,underscoring the transformative po-tential of our contributions.This work represents a significant step forward in the application of machine learning tech-niques to critical societal challenges.
      原文链接:
    • 万方数据
    • 维普

    FedBone:Towards Large-Scale Federated Multi-Task Learning

    陈益强张腾蒋鑫龙陈前...
    1040-1057页
    查看更多>>摘要：Federated multi-task learning(FMTL)has emerged as a promising framework for learning multiple tasks si-multaneously with client-aware personalized models.While the majority of studies have focused on dealing with the non-independent and identically distributed(Non-IID)characteristics of client datasets,the issue of task heterogeneity has largely been overlooked.Dealing with task heterogeneity often requires complex models,making it impractical for federat-ed learning in resource-constrained environments.In addition,the varying nature of these heterogeneous tasks introduces inductive biases,leading to interference during aggregation and potentially resulting in biased global models.To address these issues,we propose a hierarchical FMTL framework,referred to as FedBone,to facilitate the construction of large-scale models with improved generalization.FedBone leverages server-client split learning and gradient projection to split the entire model into two components:1)a large-scale general model(referred to as the general model)on the cloud serv-er,and 2)multiple task-specific models(referred to as client models)on edge clients,accommodating devices with limited compute power.To enhance the robustness of the large-scale general model,we incorporate the conflicting gradient projec-tion technique into FedBone to rectify the skewed gradient direction caused by aggregating gradients from heterogeneous tasks.The proposed FedBone framework is evaluated on three benchmark datasets and one real ophthalmic dataset.The comprehensive experiments demonstrate that FedBone efficiently adapts to the heterogeneous local tasks of each client and outperforms existing federated learning algorithms in various dense prediction and classification tasks while utilizing off-the-shelf computational resources on the client side.
      原文链接:
    • 万方数据
    • 维普

    Meta-Learning Based Few-Shot Link Prediction for Emerging Knowledge Graph

    张钰峰陈伟赵朋朋许佳捷...
    1058-1077页
    查看更多>>摘要：Inductive knowledge graph embedding(KGE)aims to embed unseen entities in emerging knowledge graphs(KGs).The major recent studies of inductive KGE embed unseen entities by aggregating information from their neighbor-ing entities and relations with graph neural networks(GNNs).However,these methods rely on the existing neighbors of unseen entities and suffer from two common problems:data sparsity and feature smoothing.Firstly,the data sparsity problem means unseen entities usually emerge with few triplets containing insufficient information.Secondly,the effective-ness of the features extracted from original KGs will degrade when repeatedly propagating these features to represent un-seen entities in emerging KGs,which is termed feature smoothing problem.To tackle the two problems,we propose a nov-el model entitled Meta-Learning Based Memory Graph Convolutional Network(MMGCN)consisting of three different components:1)the two-layer information transforming module(TITM)developed to effectively transform information from original KGs to emerging KGs;2)the hyper-relation feature initializing module(HFIM)proposed to extract type-lev-el features shared between KGs and obtain a coarse-grained representation for each entity with these features;and 3)the meta-learning training module(MTM)designed to simulate the few-shot emerging KGs and train the model in a meta-learning framework.The extensive experiments conducted on the few-shot link prediction task for emerging KGs demon-strate the superiority of our proposed model MMGCN compared with state-of-the-art methods.
      原文链接:
    • 万方数据
    • 维普

    Combining Innovative CVTNet and Regularization Loss for Robust Adversarial Defense

    王卫东李智张丽
    1078-1093页
    查看更多>>摘要：Deep neural networks(DNNs)are vulnerable to elaborately crafted and imperceptible adversarial perturba-tions.With the continuous development of adversarial attack methods,existing defense algorithms can no longer defend against them proficiently.Meanwhile,numerous studies have shown that vision transformer(ViT)has stronger robustness and generalization performance than the convolutional neural network(CNN)in various domains.Moreover,because the standard denoiser is subject to the error amplification effect,the prediction network cannot correctly classify all recon-struction examples.Firstly,this paper proposes a defense network(CVTNet)that combines CNNs and ViTs that is ap-pended in front of the prediction network.CVTNet can effectively eliminate adversarial perturbations and maintain high robustness.Furthermore,this paper proposes a regularization loss(LCPL),which optimizes the CVTNet by computing dif-ferent losses for the correct prediction set(CPS)and the wrong prediction set(WPS)of the reconstruction examples,re-spectively.The evaluation results on several standard benchmark datasets show that CVTNet performs better robustness than other advanced methods.Compared with state-of-the-art algorithms,the proposed CVTNet defense improves the av-erage accuracy of pixel-constrained attack examples generated on the CIFAR-10 dataset by 24.25％and spatially-con-strained attack examples by 14.06％.Moreover,CVTNet shows excellent generalizability in cross-model protection.
      原文链接:
    • 万方数据
    • 维普

    Sequential Cooperative Distillation for Imbalanced Multi-Task Learning

    冯泉姚佳雨谢明昆黄圣君...
    1094-1106页
    查看更多>>摘要：Multi-task learning(MTL)can boost the performance of individual tasks by mutual learning among multi-ple related tasks.However,when these tasks assume diverse complexities,their corresponding losses involved in the MTL objective inevitably compete with each other and ultimately make the learning biased towards simple tasks rather than complex ones.To address this imbalanced learning problem,we propose a novel MTL method that can equip multiple ex-isting deep MTL model architectures with a sequential cooperative distillation(SCD)module.Specifically,we first intro-duce an efficient mechanism to measure the similarity between tasks,and group similar tasks into the same block to allow their cooperative learning from each other.Based on this,the grouped task blocks are sorted in a queue to determine the learning sequence of the tasks according to their complexities estimated with the defined performance indicator.Finally,a distillation between the individual task-specific models and the MTL model is performed block by block from complex to simple manner,achieving a balance between competition and cooperation among learning multiple tasks.Extensive experi-ments demonstrate that our method is significantly more competitive compared with state-of-the-art methods,ranking No.1 with average performances across multiple datasets by improving 12.95％and 3.72％compared with OMTL and MTLKD,respectively.
      原文链接:
    • 万方数据
    • 维普

    Spatio-Temporal Learning for Route-Based Travel Time Estimation

    房子荃孙琦晨陈璐胡丹蕾...
    1107-1122页
    查看更多>>摘要：Travel time estimation(TTE)is a fundamental task to build intelligent transportation systems.However,most existing TTE solutions design models upon simple homogeneous graphs and ignore the heterogeneity of traffic net-works,where,e.g.,main roads typically contribute differently from side roads.In terms of spatial dimension,few studies consider the dynamic spatial correlations across road segments,e.g.,the traffic speed/volume on road segment A may cor-relate with the traffic speed/volume on road segment B,where A and B could be adjacent or non-adjacent,and such corre-lations may vary across time.In terms of temporal dimension,even fewer studies consider the dynamic temporal depen-dences,where,e.g.,the historical states of road A may directly correlate with the recent state of A,and may also indirect-ly correlate with the recent state of road B.To track all aforementioned issues of existing TTE approaches,we provide HDTTE,a solution that employs heterogeneous and dynamic spatio-temporal predictive learning.Specifically,we first de-sign a general multi-relational graph constructor that extracts hidden heterogeneous information of road segments,where we model road segments as nodes and model correlations as edges in the multi-relational graph.Next,we propose a dy-namic graph attention convolution module that aggregates dynamic spatial dependence of neighbor roads to focal roads.We also present a novel correlation-augmented temporal convolution module to capture the influence of states at past time steps on current traffic states.Finally,in view of the periodic dependence of traffic,we develop a multi-scale adaptive fu-sion layer to enable HDTTE to exploit periodic patterns from recent,daily,and weekly traffic states.An experimental study using real-life highway and urban datasets demonstrates the validity of the approach and its advantage over others.
      原文链接:
    • 万方数据
    • 维普

    Enhancing Recommendation with Denoising Auxiliary Task

    刘鹏圣郑力南陈加乐张广发...
    1123-1137页
    查看更多>>摘要：The historical interaction sequences of users play a crucial role in training recommender systems that can ac-curately predict user preferences.However,due to the arbitrariness of user behaviors,the presence of noise in these se-quences poses a challenge to predicting their next actions in recommender systems.To address this issue,our motivation is based on the observation that training noisy sequences and clean sequences(sequences without noise)with equal weights can impact the performance of the model.We propose the novel self-supervised Auxiliary Task Joint Training(ATJT)method aimed at more accurately reweighting noisy sequences in recommender systems.Specifically,we strategically se-lect subsets from users'original sequences and perform random replacements to generate artificially replaced noisy se-quences.Subsequently,we perform joint training on these artificially replaced noisy sequences and the original sequences.Through effective reweighting,we incorporate the training results of the noise recognition model into the recommender model.We evaluate our method on three datasets using a consistent base model.Experimental results demonstrate the ef-fectiveness of introducing the self-supervised auxiliary task to enhance the base model's performance.
      原文链接:
    • 万方数据
    • 维普

    Intent-Aware Graph-Level Embedding Learning Based Recommendation

    郝鹏翼刘思浩白琮
    1138-1152页
    查看更多>>摘要：Recommendation has been widely used in business scenarios to provide users with personalized and accurate item lists by efficiently analyzing complex user-item interactions.However,existing recommendation methods have signifi-cant shortcomings in capturing the dynamic preference changes of users and discovering their true potential intents.To address these problems,a novel framework named Intent-Aware Graph-Level Embedding Learning(IaGEL)is proposed for recommendation.In this framework,the potential user interest is explored by capturing the co-occurrence of items in different periods,and then user interest is further improved based on an adaptive aggregation algorithm,forming generic intents and specific intents.In addition,for better representing the intents,graph-level embedding learning is designed based on the mutual information comparison among positive intents and negative intents.Finally,an intent-based recom-mendation strategy is designed to further mine the dynamic changes in user preferences.Experiments on three public and industrial datasets demonstrate the effectiveness of the proposed IaGEL in the task of recommendation.
      原文链接:
    • 万方数据
    • 维普