检索结果-内蒙古大学图书馆

Reverse-Nearest-Neighbor-Based Clustering by Fast Search and Find of Density Peaks

Chinese Journal of Electronics 2023年第6期32卷 1341-1354页

作者： ZHANG Chunhao XIE Bin ZHANG Yiran College of Computer and Cyber Security Hebei Normal University Hebei Provincial Engineering Research Center for Supply Chain Big Data Analytics and Data Security Hebei Normal University Hebei Provincial Key Laboratory of Network and Information Security Hebei Normal University

Clustering by fast search and find of density peaks(CFSFDP) has the advantages of a novel idea, easy implementation, and efficient clustering. It has been widely recognized in various fields since it was proposed in Science in 2014. The CFSFDP algorithm also has certain limitations, such as non-unified sample density metrics defined by cutoff distance, the domino effect for the assignment of remaining samples triggered by unstable assignment strategy, and the phenomenon of picking wrong density peaks as cluster centers. We propose reverse-nearest-neighbor-based clustering by fast search and find of density peaks(RNN-CFSFDP) to avoid these shortcomings. We redesign and unify the sample density metric by introducing reverse nearest neighbor. The newly defined local density metric and the K-nearest neighbors of each sample are combined to make the assignment process more robust and alleviate the domino effect. A cluster fusion algorithm is proposed, which further alleviates the domino effect and effectively avoids the phenomenon of picking wrong density peaks as cluster centers. Experimental results on publicly available synthetic data sets and real-world data sets show that in most cases, the proposed algorithm is superior to or at least equivalent to the comparative methods in clustering performance. The proposed algorithm works better on manifold data sets and uneven density data sets.

关键词： Measurement Manifolds Shape Clustering algorithms Synthetic data

来源：评论

学校读者我要写书评

暂无评论

A Traceability Method for Bitcoin Transactions Based on Gateway Network Traffic Analysis

A Traceability Method for Bitcoin Transactions Based on Gate...

引用

2022 International Conference on Networking and Network Applications, NaNA 2022

作者： Huang, Dapeng Chen, Haoran Wang, Kai Chen, Chen Han, Weili Fudan University Laboratory for Data Analytics and Security China

ISBN: (数字)9781665461313

ISBN: (纸本)9781665461313

Cryptocurrencies like Bitcoin have become a popular weapon for illegal activities. They have the characteristics of decentralization and anonymity, which can effectively avoid the supervision of government departments. How to de-anonymize Bitcoin transactions is a crucial issue for regulatory and judicial investigation departments to supervise and combat crimes involving Bitcoin effectively. This paper aims to de-anonymize Bitcoin transactions and present a Bitcoin transaction traceability method based on Bitcoin network traffic analysis. According to the characteristics of the physical network that the Bitcoin network relies on, the Bitcoin network traffic is obtained at the physical convergence point of the local Bitcoin network. By analyzing the collected network traffic data, we realize the traceability of the input address of Bitcoin transactions and test the scheme in the distributed Bitcoin network environment. The experimental results show that this traceability mechanism is suitable for nodes connected to the Bitcoin network (except for VPN, Tor, etc.), and can obtain 47.5% recall rate and 70.4% precision rate, which are promising in practice. © 2022 IEEE.

关键词： Bitcoin

来源：评论

学校读者我要写书评

暂无评论

Crop Classification Methods Based on Siamese CBMM-CNN Architectures Using Hyperspectral Remote Sensing data 24th

Crop Classification Methods Based on Siamese CBMM-CNN Archi...

引用

24th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP 2024

作者： Xie, Bin Zhang, Jiahao Li, Yuling Li, Yusong Dong, Xinyu College of Computer and Cyber Security Hebei Normal University Hebei Shijiazhuang China Hebei Provincial Key Laboratory of Network and Information Security Hebei Normal University Hebei Shijiazhuang China Hebei Provincial Engineering Research Center for Supply Chain Big Data Analytics and Data Security Hebei Normal University Hebei Shijiazhuang China

ISBN: (纸本)9789819615278

Amidst global population growth and escalating food demands, real-time agricultural monitoring is crucial for ensuring food security. During the initial stages of crop growth, however, it faces significant challenges in obtaining accurate ground truth data via remote sensing, thereby impeding effective crop classification. Traditional Convolutional Neural Network (CNN) models, although effective in image processing and feature extraction within computer vision, are notably limited in their ability to handle the high dimensionality and complex spectral information inherent in hyperspectral non-image remote sensing data. To address this problem, we designed an enhanced CNN model, CBM-CNN. The CBM-CNN incorporates additional convolutional layers (C), batch normalization (B), and MLP(M) to convert one-dimensional data into two-dimensional data for classification, thus improving the processing efficiency of highly nonlinear data features. Building on CBM-CNN, we add multiple attention mechanisms (CBMM-CNN) to enhance feature extraction capabilities. The Siamese CBMM-CNN (SCBMM-CNN) is employed for in-depth learning of data features, with supervised learning precisely adjusting the model to improve classification performance. Experimental results on hyperspectral remote sensing datasets demonstrate that SCBMM-CNN outperforms traditional machine learning models (PCA-RF, GBDT, GMO-SVM, Random Forest, Decision Tree) and deep learning models (CNN, CNN-LSTM) in terms of accuracy, recall rate, F1 score, and kappa coefficient. These results confirm that SCBMM-CNN significantly improves crop classification accuracy and provides a novel method for early crop growth classification. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Private, Efficient, and Accurate: Protecting Models Trained by Multi-party Learning with Differential Privacy

Private, Efficient, and Accurate: Protecting Models Trained ...

引用

IEEE Symposium on security and Privacy

作者： Wenqiang Ruan Mingxin Xu Wenjing Fang Li Wang Lei Wang Weili Han Laboratory of Data Analytics and Security Fudan University Ant Group

Secure multi-party computation-based machine learning, referred to as multi-party learning (MPL for short), has become an important technology to utilize data from multiple parties with privacy preservation. While MPL provides rigorous security guarantees for the computation process, the models trained by MPL are still vulnerable to attacks that solely depend on access to the models. Differential privacy could help to defend against such attacks. However, the accuracy loss brought by differential privacy and the huge communication overhead of secure multi-party computation protocols make it highly challenging to balance the 3-way trade-off between privacy, efficiency, and *** this paper, we are motivated to resolve the above issue by proposing a solution, referred to as PEA (Private, Efficient, Accurate), which consists of a secure differentially private stochastic gradient descent (DPSGD for short) protocol and two optimization methods. First, we propose a secure DPSGD protocol to enforce DPSGD, which is a popular differentially private machine learning algorithm, in secret sharing-based MPL frameworks. Second, to reduce the accuracy loss led by differential privacy noise and the huge communication overhead of MPL, we propose two optimization methods for the training process of MPL: (1) the data-independent feature extraction method, which aims to simplify the trained model structure; (2) the local data-based global model initialization method, which aims to speed up the convergence of the model training. We implement PEA in two open-source MPL frameworks: TF-Encrypted and Queqiao. The experimental results on various datasets demonstrate the efficiency and effectiveness of PEA. E.g. when ϵ = 2, we can train a differentially private classification model with an accuracy of 88% for CIFAR-10 within 7 minutes under the LAN setting. This result significantly outperforms the one from CryptGPU, one state-of-the-art MPL framework: it costs more than 16 hours to tra

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Fault Diagnostic Method for Rolling Bearings Based on CAR-SNN Model

A Fault Diagnostic Method for Rolling Bearings Based on CAR-...

引用

2023 CAA Symposium on Fault Detection, Supervision and Safety for Technical Processes, SAFEPROCESS 2023

作者： Li, Qiuting Wang, Xiuqing Du, Wenxia Hebei Normal University College of Computer and Cyber Security Shijiazhuang050024 China Hebei Provincial Key Laboratory of Network & Information Security Shijiazhuang China Hebei Provincial Engineering Research Center for Supply Chain Big Data Analytics & Data Security Shijiazhuang China Hebei Normal University College of China Combustion Engineering Shijiazhuang050024 China

ISBN: (纸本)9798350337754

In the era of big data, high-performance rolling bearing fault diagnostic methods are of great significance to ensure manufacturing smoothly and avoid mass economic losses. A rolling bearing fault diagnostic CAR-SNN model based on Convolutional Block Attention Module (CBAM), Residual Network (ResNet), and Spiking Neural Networks (SNNs) is proposed based on deep learning (DL). The CAR-SNN model combines CBAM, ResNet, and SNNs to extract critical features while avoiding gradient vanishing or explosion phenomena. The ability of the CAR-SNN model to process spatio-temporal information is improved by using SNNs that incorporate spatio-temporal information simultaneously. The proposed model is used for the rolling bearing fault diagnosis, which datasets are from the experimental platforms of CWRU and Jiangnan University, with the highest diagnostic accuracy 99.995% and 99.18%, respectively. The experimental results validate the effectiveness and stability of the proposed model. The results of comparison experiments show that the CAR-SNN model has better fault diagnostic accuracy and stability than that of the CAR model and other comparison methods. The proposed data-driven rolling bearing fault diagnostic method based on the CAR-SNN model is easy to be implemented, and meeting the needs of high fault diagnostic accuracy in practical applications, and provides valuable input for fault diagnosis of rolling bearings. © 2023 IEEE.

关键词： Failure analysis

来源：评论

学校读者我要写书评

暂无评论

Improving real-world password guessing attacks via bi-directional transformers 23

Improving real-world password guessing attacks via bi-direct...

引用

Proceedings of the 32nd USENIX Conference on security Symposium

作者： Ming Xu Jitao Yu Xinyi Zhang Chuanwang Wang Shenghao Zhang Haoqi Wu Weili Han Laboratory for Data Analytics and Security Fudan University Applied AI Research Facebook

ISBN: (纸本)9781939133373

Password guessing attacks, prevalent issues in the real world, can be conceptualized as efforts to approximate the probability distribution of text tokens. Techniques in the natural language processing (NLP) field naturally lend themselves to password guessing. Among them, bi-directional transformers stand out with their ability to utilize bi-directional contexts to capture the nuances in *** further improve password guessing attacks, we propose a bi-directional-transformer-based guessing framework, referred to as PassBERT, which applies the pre-training / finetuning paradigm to password guessing attacks. We first prepare a pre-trained password model, which contains the knowledge of the general password distribution. Then, we design three attack-specific fine-tuning approaches to tailor the pretrained password model to the following real-world attack scenarios: (1) conditional password guessing, which recovers the complete password given a partial password; (2) targeted password guessing, which compromises the password(s) of a specific user using their personal information; (3) adaptive rule-based password guessing, which selects adaptive mangling rules for a word (i.e., base password) to generate rule-transformed password candidates. The experimental results show that our fine-tuned models can outperform the state-of-the-art models by 14.53%, 21.82% and 4.86% in the three attacks, respectively, demonstrating the effectiveness of bi-directional transformers on downstream guessing attacks. Finally, we propose a hybrid password strength meter to mitigate the risks from the three attacks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Interactive Attention Mechanism Fusion Network for Aspect-Based Multimodal Sentiment Analysis

An Interactive Attention Mechanism Fusion Network for Aspect...

引用

2023 International Conference on Machine Learning and Cybernetics, ICMLC 2023

作者： Wang, Jinghong Gao, Yuan Li, Haokang College of Computer and Cyber Security Hebei Normal University Shijiazhuang050024 China Hebei Normal University Hebei Key Laboratory of Network and Information Security Shijiazhuang050024 China Hebei Normal University Hebei Provincial Engineering Research Center for Supply Chain Big Data Analytics & Data Security Shijiazhuang050024 China Hebei Polytechnic Institute College of Artificial Intelligence and Big Data Shijiazhuang050000 China

ISBN: (纸本)9798350303780

The goal of aspect-based multimodal sentiment analysis (ABMSA) is to classify the sentiment associated with aspect words in a given context. Most current ABMSA models focus only on general inter-modal information interactions without considering both intra-modal and inter-modal information interactions and ignoring image noise. To address these issues, this paper proposes an Interactive Attention Mechanism Fusion Network (IAMFN) model. The model first designs an image-text fusion module based on the attention, which applies the attention mechanism to a recurrent neural network to fuse text and images while filtering the noise in the images, and finally adds the fused information to the aspect information step by step to obtain the dynamic inter-modal representation. In addition, this paper proposes an aspect-text fusion module based on the attention, which learns the intra-modal contextual representation by calculating the weights of each aspect word in the context. Finally, this paper stitches the information obtained from the two modules and feeds it into the fully connected softmax layers to predict sentiment polarity. We have conducted extensive experiments on two benchmark datasets and the experimental results show that our model achieves state-of-the-art performance. © 2023 IEEE.

关键词： Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

Template-Order Driven Feature Integration With Generative Models for Aspect Sentiment Triplet Extraction

IEEE Transactions on Audio, Speech and Language Processing

引用

IEEE Transactions on Audio, Speech and Language Processing 2025年 33卷 2275-2285页

作者： Jiazhou Chen Ruiqiang Guo College of Computer and Cyber Security Hebei Normal University Shijiazhuang China Hebei Provincial Key Laboratory of Network & Information Security Hebei Normal University Shijiazhuang China Hebei Provincial Engineering Research Center for Supply Chain Big Data Analytics & Data Security Hebei Normal University Shijiazhuang China

Aspect-based sentiment analysis (ABSA), which explores the nuanced sentiments individuals express toward specific services or products, has shown significant potential in practical applications. Recently, the aspect sentiment triplet extraction field has witnessed significant advancements through the prowess of generative models. However, existing methods face critical limitations: (1) fixed-order decoding (e.g., aspect term $\rightarrow$ opinion term $\rightarrow$ sentiment polarity) ignores interdependencies between elements; (2) token-by-token generation fails to model term boundaries, which struggles with multi-word terms or sentences containing multiple triplets. To address these challenges, we propose a Template-Order driven Feature Integration (TOFI) framework, which integrates two novel modules: Template-Order Prompt (TOP) and Feature Progressive Sequence Labeling (FPSL). TOP employs various orders of sentiment elements to prompt the model to generate sentiment tuples from different perspectives, each using a different element order. It then aggregates the sentiment tuples that appear across all orders. FPSL leverages sequence labeling to integrate the rich semantic information of aspect terms and opinion terms into the model, enhancing its ability to capture semantic nuances. Furthermore, the TOFI framework introduces unified multi-domain training, prefixing domain identifiers to inputs for multi-task learning across datasets. Extensive experiments on multiple public benchmarks demonstrate that the proposed framework consistently surpassed representative baseline models.

关键词： Labeling Decoding Training Feature extraction Speech processing Semantics Sentiment analysis Pipelines data mining Benchmark testing

来源：评论

学校读者我要写书评

暂无评论

The Connections Between the Approximate Three-Way Concept Lattice and the Three-Way Approximate Concept Lattice 11th

The Connections Between the Approximate Three-Way Concept L...

引用

11th International Joint Conference on Rough Sets, IJCRS 2025

作者： Li, Mei-Zheng Li, Lei-Jun Mi, Ju-Sheng Hebei Normal University Hebei Shijiazhuang050024 China Hebei Provincial Engineering Research Center for Supply Chain Big Data Analytics and Data Security Shijiazhuang050024 China Hebei Key Laboratory of Computational Mathematics and Applications Shijiazhuang050024 China

ISBN: (纸本)9783031927430

The approximate three-way concept lattice (ATWL), a novel model for conflict analysis, is an extension of classical concept lattice in formal concept analysis. However, little effort has been devoted on the connections between the ATWL and other existing three-way concept lattice models, which will be of great help to construct the ATWL efficiently. This paper reveals the connections between the ATWL and the three-way approximate concept lattice (TWAL) of an incomplete formal context from the viewpoints of concept fusion. Furthermore, the construct algorithm for ATWL based on TWAL is presented. The obtained results are finally demonstrated and verified by an example. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Approximate concept construction Concept lattice Incomplete formal context Three-way decisions

来源：评论

学校读者我要写书评

暂无评论

PII-Bench: Evaluating Query-Aware Privacy Protection Systems

arXiv

引用

arXiv 2025年

作者： Shen, Hao Gu, Zhouhong Hong, Haokai Han, Weili Institute of Fintech Fudan University China Shanghai Key Laboratory of Data Science School of Computer Science Fudan University China Laboratory of Data Analytics and Security Fudan University China

The widespread adoption of Large Language Models (LLMs) has raised significant privacy concerns regarding the exposure of personally identifiable information (PII) in user prompts. To address this challenge, we propose a query-unrelated PII masking strategy and introduce PII-Bench, the first comprehensive evaluation framework for assessing privacy protection systems. PII-Bench comprises 2,842 test samples across 55 fine-grained PII categories, featuring diverse scenarios from single-subject descriptions to complex multi-party interactions. Each sample is carefully crafted with a user query, context description, and standard answer indicating query-relevant PII. Our empirical evaluation reveals that while current models perform adequately in basic PII detection, they show significant limitations in determining PII query relevance. Even state-of-the-art LLMs struggle with this task, particularly in handling complex multi-subject scenarios, indicating substantial room for improvement in achieving intelligent PII masking. © 2025, CC BY.

关键词： Differential privacy

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：