检索结果-内蒙古大学图书馆

On large language models safety,security,and privacy:A survey

Journal of Electronic science and Technology 2025年第1期23卷 1-21页

作者： Ran Zhang Hong-Wei Li Xin-Yuan Qian Wen-Bo Jiang Han-Xiao Chen School of Computer Science and Engineering University of Electronic Science and Technology of ChinaChengdu611731China

The integration of artificial intelligence(AI)technology,particularly large language models(LLMs),has become essential across various sectors due to their advanced language comprehension and generation *** their transformative impact in fields such as machine translation and intelligent dialogue systems,LLMs face significant *** challenges include safety,security,and privacy concerns that undermine their trustworthiness and effectiveness,such as hallucinations,backdoor attacks,and privacy *** works often conflated safety issues with security *** contrast,our study provides clearer and more reasonable definitions for safety,security,and privacy within the context of *** on these definitions,we provide a comprehensive overview of the vulnerabilities and defense mechanisms related to safety,security,and privacy in ***,we explore the unique research challenges posed by LLMs and suggest potential avenues for future research,aiming to enhance the robustness and reliability of LLMs in the face of emerging threats.

关键词： Large language models Privacy issues Safety issues Security issues

来源：评论

学校读者我要写书评

暂无评论

DyHDGE:Dynamic heterogeneous transaction graph embedding for safety-centric fraud detection in financial scenarios

引用

Journal of Safety science and Resilience 2024年第4期5卷 486-497页

作者： Xinzhi Wang Jiayu Guo Xiangfeng Luo Hang Yu School of Computer Engineering and Science Shanghai UniversityShanghai200444China

Dynamic graph fraud detection aims to distinguish fraudulent entities that deviate significantly from most benign entities within an ever-changing graph ***,when dealing with different financial fraud scenarios,existing methods face challenges,resulting in difficulty in effectively ensuring financial *** fraud scenarios,transaction data are generated in real time,in which a strong temporal relationship between multiple fraudulent transactions is *** dynamic graph models struggle to effectively balance the temporal features of nodes and spatial structural features,failing to handle different types of nodes in the graph *** this study,to extract the temporal and structural information,we proposed a dynamic heterogeneous transaction graph embedding(DyHDGE)network based on a dynamic heterogeneous transaction graph,considering both temporal and structural information while incorporating heterogeneous *** separately extract temporal relationships between transactions and spatial structural relationships between nodes,we used a heterogeneous temporal graph representation learning module and a temporal graph structure information extraction ***,we designed two loss functions to optimize node feature *** experiments demonstrated that the proposed DyHDGE significantly outperformed previous state-of-the-art methods on two simulated datasets of financial fraud *** capability contributes to enhancing security in financial consumption scenarios.

关键词： Dynamic graph learning Fraud detection Graph neural network Temporal information

来源：评论

学校读者我要写书评

暂无评论

Multi-Scale Feature Optimization Point Cloud Completion Network Integrating SoftPool

IAENG International Journal of Computer Science

引用

IAENG International Journal of computer science 2025年第1期52卷 233-243页

作者： Zhang, Wanpeng Zhou, Ziwei School of Computer Science and Software Engineering University of Science and Technology LiaoNing Anshan114051 China School of Computer Science and Software Engineering University of Science and Technology LiaoNing Anshan114051 China

Point cloud completion is crucial in point cloud processing, as it can repair and refine incomplete 3D data, ensuring more accurate models. However, current point cloud completion methods commonly face a challenge: they fail to fully utilize multi-scale information from local features, leading to limitations in accuracy and detail preservation. To address this issue, this paper proposes a multi-scale feature optimization algorithm for point cloud completion that integrates SoftPool. Based on DGCNN, the method combines dilated convolution and bottleneck attention mechanisms to extract features at different scales, enhancing the ability to capture detailed information in point clouds. The bottleneck attention mechanism is used to optimize important detail features. The extracted local features are concatenated with their corresponding positional information to form point proxies, enhancing the effective extraction of local geometric features, resulting in more refined completed point cloud shapes. A Transformer architecture is employed to model these features. Finally, SoftPool is introduced for fine-grained feature downsampling, improving the network's ability to recover point cloud details. FoldingNet is used to reconstruct missing structures and output the completed point cloud. To validate the model's completion performance, training and testing are conducted on the PCN and ShapeNet55 datasets. Experimental results demonstrate that the model has better feature detail retention and more accurate completion results. On the PCN dataset, the average CD value is reduced by 6.5% compared to the best-performing model among the comparison methods. On the ShapeNet55 dataset, the average CD value across three difficulty levels is reduced by 6.9% compared to the best-performing model among the comparison methods. Additionally, the model also achieved a 2.1% improvement in F-score. © (2025), (International Association of Engineers). All rights reserved.

关键词： Ability testing

来源：评论

学校读者我要写书评

暂无评论

A survey on cross-user federated recommendation

引用

science China(Information sciences) 2025年第4期68卷 7-32页

作者： Enyue YANG Yudi XIONG Wei YUAN Weike PAN Qiang YANG Zhong MING College of Computer Science and Software Engineering Shenzhen University School of Electrical Engineering and Computer Science The University of Queensland WeBank AI Lab WeBank Department of Computer Science and Engineering Hong Kong University of Science and Technology College of Big Data and Internet Shenzhen Technology University Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ)

Recommender systems are effective in mitigating information overload, yet the centralized storage of user data raises significant privacy concerns. Cross-user federated recommendation(CUFR) provides a promising distributed paradigm to address these concerns by enabling privacy-preserving recommendations directly on user devices. In this survey, we review and categorize current progress in CUFR, focusing on four key aspects: privacy, security, accuracy, and efficiency. Firstly,we conduct an in-depth privacy analysis, discuss various cases of privacy leakage, and then review recent methods for privacy protection. Secondly, we analyze security concerns and review recent methods for untargeted and targeted *** untargeted attack methods, we categorize them into data poisoning attack methods and parameter poisoning attack methods. For targeted attack methods, we categorize them into user-based methods and item-based methods. Thirdly,we provide an overview of the federated variants of some representative methods, and then review the recent methods for improving accuracy from two categories: data heterogeneity and high-order information. Fourthly, we review recent methods for improving training efficiency from two categories: client sampling and model compression. Finally, we conclude this survey and explore some potential future research topics in CUFR.

关键词： cross-user federated recommendation federated recommendation federated learning recommender systems user privacy

来源：评论

学校读者我要写书评

暂无评论

Improving UAV Image Target Detection: A Novel Approach Using OptiDETR with Swin Transformer

IAENG International Journal of Computer Science

引用

IAENG International Journal of computer science 2025年第3期52卷 771-780页

作者： Ma, Wenlong Liu, Weisheng School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan114051 China School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan114051 China

In the analysis of drone aerial images, object detection tasks are particularly challenging, especially in the presence of complex terrain structures, extreme differences in target sizes, suboptimal shooting angles, and varying lighting conditions, all of which exacerbate the difficulty of recognition. In recent years, the DETR model based on the Transformer architecture has eliminated traditional post-processing steps such as NMS(Non-Maximum Suppression), thereby simplifying the object detection process and improving detection accuracy, which has garnered widespread attention in the academic community. However, DETR has limitations such as slow training convergence, difficulty in query optimization, and high computational costs, which hinder its application in practical fields. To address these issues, this paper proposes a new object detection model called OptiDETR. This model first employs a more efficient hybrid encoder to replace the traditional Transformer encoder. The new encoder significantly enhances feature processing capabilities through internal and cross-scale feature interaction and fusion logic. Secondly, an IoU (Intersection over Union) aware query selection mechanism is introduced. This mechanism adds IoU constraints during the training phase to provide higher-quality initial object queries for the decoder, significantly improving the decoding performance. Additionally, the OptiDETR model integrates SW-Block into the DETR decoder, leveraging the advantages of Swin Transformer in global context modeling and feature representation to further enhance the performance and efficiency of object detection. To tackle the problem of small object detection, this study innovatively employs the SAHI algorithm for data augmentation. Through a series of experiments, It achieved a significant performance improvement of more than two percentage points in the mAP (mean Average Precision) metric compared to current mainstream object detection models. Furthermore, ther

关键词： Decoding

来源：评论

学校读者我要写书评

暂无评论

A privacy-preserving method for publishing data with multiple sensitive attributes

引用

CAAI Transactions on Intelligence Technology 2024年第1期9卷 222-238页

作者： Tong Yi Minyong Shi Wenqian Shang Haibin Zhu School of Computer Science and Engineering Guangxi Normal UniversityGuilinChina School of Computer Science Communication University of ChinaBeijingChina Department of Computer Science and Mathematics Nipissing UniversityNorth BayOntarioCanada

The overgeneralisation may happen because most studies on data publishing for multiple sensitive attributes(SAs)have not considered the personalised privacy ***,sensitive information disclosure may also be caused by these personalised *** address the matter,this article develops a personalised data publishing method for multiple *** to the requirements of individuals,the new method partitions SAs values into two categories:private values and public values,and breaks the association between them for privacy *** the private values,this paper takes the process of anonymisation,while the public values are released without this *** algorithm is designed to achieve the privacy mode,where the selectivity is determined by the sensitive value frequency and undesirable *** experimental results show that the proposed method can provide more information utility when compared with previous *** theoretic analyses and experiments also indicate that the privacy can be guaranteed even though the public values are known to an *** overgeneralisation and privacy breach caused by the personalised requirement can be avoided by the new method.

关键词： data privacy data publishing

来源：评论

学校读者我要写书评

暂无评论

Deep learning-based software engineering: progress,challenges, and opportunities

引用

science China(Information sciences) 2025年第1期68卷 57-144页

作者： Xiangping CHEN Xing HU Yuan HUANG He JIANG Weixing JI Yanjie JIANG Yanyan JIANG Bo LIU Hui LIU Xiaochen LI Xiaoli LIAN Guozhu MENG Xin PENG Hailong SUN Lin SHI Bo WANG Chong WANG Jiayi WANG Tiantian WANG Jifeng XUAN Xin XIA Yibiao YANG Yixin YANG Li ZHANG Yuming ZHOU Lu ZHANG School of Journalism and Communication Sun Yat-sen University School of Software Technology Zhejiang University School of Software Engineering Sun Yat-sen University School of Software Dalian University of Technology School of Computer Science and Technology Beijing Institute of Technology Key Laboratory of High Confidence Software Technologies (Peking University) Ministry of EducationSchool of Computer Science Peking University State Key Laboratory for Novel Software Technology Nanjing University School of Computer Science and Engineering Beihang University Institute of Information Engineering Chinese Academy of Sciences School of Computer Science Fudan University State Key Laboratory of Complex & Critical Software Environment (CCSE) School of Software Beihang University School of Computer and Information Technology Beijing Jiaotong University School of Computer Science and Technology Harbin Institute of Technology School of Computer Science Wuhan University Huawei Technologies

Researchers have recently achieved significant advances in deep learning techniques, which in turn has substantially advanced other research disciplines, such as natural language processing, image processing, speech recognition, and software engineering. Various deep learning techniques have been successfully employed to facilitate software engineering tasks, including code generation, software refactoring, and fault localization. Many studies have also been presented in top conferences and journals, demonstrating the applications of deep learning techniques in resolving various software engineering tasks. However,although several surveys have provided overall pictures of the application of deep learning techniques in software engineering,they focus more on learning techniques, that is, what kind of deep learning techniques are employed and how deep models are trained or fine-tuned for software engineering tasks. We still lack surveys explaining the advances of subareas in software engineering driven by deep learning techniques, as well as challenges and opportunities in each subarea. To this end, in this study, we present the first task-oriented survey on deep learning-based software engineering. It covers twelve major software engineering subareas significantly impacted by deep learning techniques. Such subareas spread out through the whole lifecycle of software development and maintenance, including requirements engineering, software development, testing, maintenance, and developer collaboration. As we believe that deep learning may provide an opportunity to revolutionize the whole discipline of software engineering, providing one survey covering as many subareas as possible in software engineering can help future research push forward the frontier of deep learning-based software engineering more systematically. For each of the selected subareas,we highlight the major advances achieved by applying deep learning techniques with pointers to the available datasets i

关键词： deep learning software engineering software benchmark software artifact representation survey

来源：评论

学校读者我要写书评

暂无评论

Enabling Harmonious Human-Machine Interaction with Visual-Context Augmented Dialogue System: A Review

引用

ACM Transactions on Information Systems 2025年第3期43卷 1-59页

作者： Wang, Hao Guo, Bin Zeng, Yating Chen, Mengqi Ding, Yasan Zhang, Ying Yao, Lina Yu, Zhiwen School of Computer Science Northwestern Polytechnical University Xi'an China School of Computer Science and Engineering University of New South Wales Sydney Australia

The intelligent dialogue system, aiming at communicating with humans harmoniously with natural language, is brilliant for promoting the advancement of human-machine interaction in the era of artificial intelligence. With the gradually complex human-computer interaction requirements, it is difficult for traditional text-based dialogue system to meet the demands for more vivid and convenient interaction. Consequently, Visual-Context Augmented Dialogue (VAD) System, which has the potential to communicate with humans by perceiving and understanding multimodal information (i.e., visual context in images or videos, textual dialogue history), has become a predominant research paradigm. Benefiting from the consistency and complementarity between visual and textual context, VAD possesses the potential to generate engaging and context-aware responses. To depict the development of VAD, we first characterize the concept model of VAD and then present its generic system architecture to illustrate the system workflow, followed by a summary of multimodal fusion techniques. Subsequently, several research challenges and representative works are investigated, followed by the summary of authoritative benchmarks and real-world application of VAD. We conclude this article by putting forward some open issues and promising research trends for VAD, e.g., the cognitive mechanisms of human-machine dialogue under cross-modal dialogue context, mobile and lightweight deployment of VAD. © 2025 Copyright held by the owner/author(s).

关键词： Human computer interaction

来源：评论

学校读者我要写书评

暂无评论

Federated Class-Incremental Learning via Weighted Aggregation and Distillation

引用

IEEE Internet of Things Journal 2025年第12期12卷 22489-22503页

作者： Wu, Feng Tan, Alysa Ziying Feng, Siwei Yu, Han Deng, Tao Zhao, Libang Chen, Yuanlu Soochow University School of Computer Science and Technology Suzhou215000 China Nanyang Technological University School of Computer Science and Engineering 639798 Singapore

Federated Class-Incremental Learning (FCIL) aims to design privacy-preserving collaborative training methods to continuously learn new classes from distributed datasets. In these scenarios, federated clients face the challenge of encountering new classes while being constrained by limited memory capacity, which can lead to catastrophic forgetting in the resulting global model. Existing FCIL approaches tend to overlook the challenges posed by the heterogeneity of dataset label distribution among clients, thereby constraining the generalization capacity of the global model they learn. Some of these methods also suffer from excessive computational burdens when addressing catastrophic forgetting problems. Furthermore, certain approaches are constrained to handling only straightforward data, posing significant difficulties in managing complex datasets and tackling more intricate scenarios. In this paper, we propose the Weighted Aggregation and Distillation-based FCIL (WAD-FCIL) method to address these limitations. To address data heterogeneity arising from class imbalance, we first introduce a task-aware client clustering method to identify clients with extreme class deviations before global model aggregation to eliminate potential impact on the global model. Then, we propose a multi-sampling weighted aggregation approach during the global FL model update that integrates knowledge from different clients and dynamically adjusts the weight of each client model to facilitate model update. To mitigate catastrophic forgetting, we propose a multi-model distillation strategy that involves selecting multiple teacher models for knowledge distillation. Extensive experiments comparing WAD-FCIL with ten state-of-the-art methods demonstrate that it significantly outperforms the baselines by 0.8%-3.2% in terms of average test accuracy on three representative benchmark datasets. © 2025 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Self-supervised Hypergraph Transformer with Alignment and Uniformity for Recommendation

IAENG International Journal of Computer Science

引用

IAENG International Journal of computer science 2024年第3期51卷 292-300页

作者： Yang, XianFeng Liu, Yang School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan China School of computer Science and Software Engineering University of Science and Technology Liaoning Anshan China

Graph neural networks have proven their effectiveness for user-item interaction graph collaborative filtering. However, most of the existing recommendation models highly depended on abundant and high-quality datasets and neglected the alignment and uniformity of embedding representation. The noisy and skewed distribution of data from real-world applications is ubiquitous. The alignment and uniformity are crucial for representation learning. In this paper, we propose SHTAU (self-supervised hypergraph transformer with alignment and uniformity for the recommendation) to address the problem. Specifically, we first perform a graph neural network on the user-item interaction graph, then captures the global collaboration relationship between users and items through the hypergraph transformer structure, and use the extracted global information to generate self-supervised signals for data enhancement on the user-item interaction graph to enhance the robustness of the recommendation algorithm. Simultaneously, alignment and uniformity auxiliary task update embeddings from different angles to improve performance. Extensive experiments are conducted on Yelp, Gowalla, MovieLens, Amazon-book and Tmall datasets. The experimental results show that SHTAU has obvious advantages over the baseline methods. © (2024), (International Association of Engineers). All Rights Reserved.

关键词： Alignment

来源：评论

学校读者我要写书评

暂无评论

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

请选择保存的检索档案：

请选择收藏分类：

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

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

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：