检索结果-南通市图书馆

Few-Shot Graph Classification with Structural-Enhanced Contrastive Learning for Graph Data Copyright Protection

Tsinghua Science and Technology 2024年第2期29卷 605-616页

作者： Kainan Zhang DongMyung Shin daehee seo Zhipeng Cai Department of Computer Science Georgia State UniversityAtlantaGA 30303USA LSWare Inc. Seoul 08504Republic of Korea College of Intelligence Information Engineering Sangmyung UniversitySeoul 03016Republic of Korea

Open-source licenses can promote the development of machine learning by allowing others to access,modify,and redistribute the training ***,not all open-source licenses may be appropriate for data sharing,as some may not provide adequate protections for sensitive or personal information such as social network ***,some data may be subject to legal or regulatory restrictions that limit its sharing,regardless of the licensing model ***,obtaining large amounts of labeled data can be difficult,time-consuming,or expensive in many real-world ***-shot graph classification,as one application of meta-learning in supervised graph learning,aims to classify unseen graph types by only using a small amount of labeled ***,the current graph neural network methods lack full usage of graph structures on molecular graphs and social network *** structural features are known to correlate with molecular properties in chemistry,structure information tends to be ignored with sufficient property information ***,the common binary classification task of chemical compounds is unsuitable in the few-shot setting requiring novel ***,this paper focuses on the graph classification tasks of a social network,whose complex topology has an uncertain relationship with its nodes'*** two multi-class graph datasets with large node-attribute dimensions constructed to facilitate the research,we propose a novel learning framework that integrates both meta-learning and contrastive learning to enhance the utilization of graph topological *** experiments demonstrate the competitive performance of our framework respective to other state-of-the-art methods.

关键词： few-shot learning contrastive learning data copyright protection

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Lightweight Super-Resolution Model for Complete Model Copyright Protection

引用

Tsinghua Science and Technology 2024年第4期29卷 1194-1205页

作者： Bingyi Xie Honghui Xu YongJoon Joe daehee seo Zhipeng Cai Department of Computer Science Georgia State UniversityAtlantaGA 30303USA LSWare Inc. Seoul 08504Republic of Korea College of Intelligence Information Engineering Sangmyung UniversityRepublic of KoreaSeoul 03016

Deep learning based techniques are broadly used in various applications, which exhibit superior performance compared to traditional methods. One of the mainstream topics in computer vision is the image super-resolution task. In recent deep learning neural networks, the number of parameters in each convolution layer has increased along with more layers and feature maps, resulting in better image super-resolution performance. In today’s era, numerous service providers offer super-resolution services to users, providing them with remarkable convenience. However, the availability of open-source super-resolution services exposes service providers to the risk of copyright infringement, as the complete model could be vulnerable to leakage. Therefore, safeguarding the copyright of the complete model is a non-trivial concern. To tackle this issue, this paper presents a lightweight model as a substitute for the original complete model in image super-resolution. This research has identified smaller networks that can deliver impressive performance, while protecting the original model’s copyright. Finally, comprehensive experiments are conducted on multiple datasets to demonstrate the superiority of the proposed approach in generating super-resolution images even using lightweight neural networks.

关键词： lightweight adversarial learning image super-resolution copyright protection

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Link-Privacy Preserving Graph Embedding Data Publication with Adversarial Learning

引用

Tsinghua Science and Technology 2022年第2期27卷 244-256页

作者： Kainan Zhang Zhi Tian Zhipeng Cai daehee seo the Department of Computer Science Georgia State UniversityAtlantaGA 30303USA the Department of Electrical and Computer Engineering George Mason UniversityFairfaxVA 22030USA the National Center of Excellence in Software Sangmyung UniversitySeoul 03016Republic of Korea

The inefficient utilization of ubiquitous graph data with combinatorial structures necessitates graph embedding methods,aiming at learning a continuous vector space for the graph,which is amenable to be adopted in traditional machine learning algorithms in favor of vector *** embedding methods build an important bridge between social network analysis and data analytics,as social networks naturally generate an unprecedented volume of graph data *** social network data not only brings benefit for public health,disaster response,commercial promotion,and many other applications,but also gives birth to threats that jeopardize each individual’s privacy and ***,most existing works in publishing social graph embedding data only focus on preserving social graph structure with less attention paid to the privacy issues inherited from social *** be specific,attackers can infer the presence of a sensitive relationship between two individuals by training a predictive model with the exposed social network *** this paper,we propose a novel link-privacy preserved graph embedding framework using adversarial learning,which can reduce adversary’s prediction accuracy on sensitive links,while persevering sufficient non-sensitive information,such as graph topology and node attributes in graph *** experiments are conducted to evaluate the proposed framework using ground truth social network datasets.

关键词： graph embedding privacy preservation adversarial learning

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

A Novel Transparent and Auditable Fog-Assisted Cloud Storage with Compensation Mechanism

引用

Tsinghua Science and Technology 2020年第1期25卷 28-43页

作者： Donghyun Kim Junggab Son daehee seo Yeojin Kim Hyobin Kim Jung Taek seo Department of Computer Science Kennesaw State UniversityMariettaGA 30060USA Department of Information Security Engineering Soonchunhyang UniversityAsan 31538South Korea

This paper introduces a new fog-assisted cloud storage which can achieve much higher throughput compared to the traditional cloud-only storage architecture by reducing the traffics toward the cloud storage. The fog-storage service providers are transparency to end-users and therefore, no modification on the end-user devices is necessary. This new system is featured with(1) a stronger audit scheme which is naturally coupled with the proposed architecture and does not suffer from the replay attack and(2) a transparent and efficient compensation mechanism for the fog-storage service providers. We provide rigorous theoretical analysis on the correctness and soundness of the proposed system. To the best of our knowledge, this is the first paper to discuss about a storage data audit scheme for fog-assisted cloud storage as well as the compensation mechanism for the service providers of the fog-storage service providers.

关键词： fog computing cloud computing network storage data audit merkle hash tree integer factorization payment transparency

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

An unsupervised anomaly detection framework for detecting anomalies in real time through network system’s log files analysis

引用

High-Confidence Computing 2021年第2期1卷 1-6页

作者： Vannel Zeufack Donghyun Kim daehee seo Ahyoung Lee Department of Computer Science Kennesaw State UniversityMariettaGA 30060USA Department of Computer Science Georgia State UniversityAtlantaGA 30303USA Faculty of Artificial Intelligence and Data Engineering Sangmyung UniversitySeoul 03016South Korea

Nowadays,in almost every computer system,log files are used to keep records of occurring *** log files are then used for analyzing and debugging system *** to this important utility,researchers have worked on finding fast and efficient ways to detect anomalies in a computer system by analyzing its log *** in log-based anomaly detection can be divided into two main categories:batch log-based anomaly detection and streaming log-based anomaly *** log-based anomaly detection is computationally heavy and does not allow us to instantaneously detect *** the other hand,streaming anomaly detection allows for immediate ***,current streaming approaches are mainly *** this work,we propose a fully unsupervised framework which can detect anomalies in real *** test our framework on hdfs log files and successfully detect anomalies with an F-1 score of 83%.

关键词： Anomaly detection Unsupervised machine learning Clustering OPTICS Log analysis

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Secure verifiable aggregation for blockchain-based federated averaging

引用

High-Confidence Computing 2022年第1期2卷 54-61页

作者： Saide Zhu Ruinian Li Zhipeng Cai Donghyun Kim daehee seo Wei Li Department of Computer Science Georgia State UniversityAtlantaUSA Department of Computer Science Bowling Green State UniversityBowling GreenOhioUSA Information Security Lab Sangmyung UniversitySeoulKorea

IoT devices’storage and computation capacities are constantly increasing in recent years,which brings critical challenges in data privacy *** learning(FL)and blockchain technology are two popular tech-niques used in IoT data aggregation,where FL enables data training with privacy protection,and blockchain provides a decentralized architecture for data storage and ***,very few the state-of-the-art works consider the applicability of the combination of FL and *** this paper,we adopt the federated aver-aging algorithm to reduce the communication overhead between the blockchain and end users to achieve higher *** also apply the double-mask-then-encrypt approach for end users to submit their local updates in order to protect data ***,we propose and implement a non-interactive Public Verifiable Secret Sharing(PVSS)algorithm with Distributed Hash Table(DHT)that solves the user-drop-out problem and improves the communication efficiency between blockchain and *** last,we theoretically analyze the security strengths of the proposed solution and conduct experiments to measure the execution time of PVSS on both the server and clients sides.

关键词： Blockchain Federated learning Secret sharing Privacy

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Facile synthesis of fully ordered L10-FePt nanoparticles with controlled Pt-shell thicknesses for electrocatalysis

引用

Nano Research 2017年第8期10卷 2866-2880页

作者： Yonghoon Hong Hee Jin Kim daehee Yang Gaehang Lee Ki Min Nam Myung-Hwa Jung Young-Min Kim Sang-II Choi Won seok seo Department of Chemistry Sogang University Seou104107 Republic of Korea Department of Chemistry and Green-Nano Materials Research Center Kyungpook National University Daegu 41566 Repubfic of Korea Korea Basic Science Institute and University of Science and Technology Daejeon 34133 Republic of Korea Department of Chemistry Mokpo National University Jeonnam 58554 Republic of Korea Department of Physics Sogang University Seoul 04107 Republic of Korea Center for Integrated Nanostructure Physics Institute for Basic Science (IBS) Suwon 16419 Repubfic of Korea Department of Energy Science Sungkyunkwan University (SKKU) Suwon 16419 Republic of Korea

We report a simple one-step approach for the synthesis of -4 nm uniform and fully L10-ordered face-centered tetragonal （fct） FePt nanoparticles （NPs） embedded in -60 nm MCM-41 （fct-FePt NPs@MCM-41）. We controlled the Pt-shell thickness of the fct-FePt NPs by treating the fct-FePt NPs@MCM-41 with acetic acid （HOAc） or hydrochloric acid （HC1） under sonicafion, thereby etching the surface Fe atoms of the NPs. The fct-FePt NPs deposited onto the carbon support （fct-FePt NP/C） were prepared by mixing the fct-FePt NPs@MCM-41 with carbon and subsequently removing the MCM-41 using NaOH. We also developed a facile method to synthesize acid-treated fct-FePt NP/C by using a HF solution for simultaneous surface-Fe etching and MCM-41 removal. We studied the effects of both surface-Fe etching and Pt-shell thickness on the electrocatalytic properties of fct-FePt NPs for the methanol oxidation reaction （MOR）. Compared with the non-treated fct-FePt NP/C catalyst, the HOAc-treated and HCl-treated catalysts exhibit up to 34% larger electrochemically active surface areas （ECASAs）; in addition, the HCl-treated fct-FePt NP （with -1.0 nm Pt shell）/C catalyst exhibits the highest specific activity. The HF-treated fct-FePt NP/C exhibits an ECASA almost 2 times larger than those of the other acid-treated fct-FePt NP/C catalysts and shows the highest mass activity （1,435 mA·mgPt^-1, 2.3 times higher than that of the commercial Pt/C catalyst） and stability among the catalysts tested. Our findings demonstrate that the surface-Fe etching for the generation of the Pt shell on fct-FePt NPs and the Pt-shell thickness can be factors for optimizing the electrocatalysis of the MOR.

关键词： FePt face centered tetragonal bimetallic methanol oxidation reaction electrocatalysts

来源：

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

欢迎您,

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：