检索结果-内蒙古大学图书馆

Anomaly Detection in Social Media Texts Using Optimal Convolutional Neural Network

Intelligent Automation & Soft Computing 2023年第4期36卷 1027-1042页

作者： Swarna Sudha Muppudathi Valarmathi Krishnasamy Department of Computer Science and Engineering Ramco Institute of TechnologyRajapalayam626117India Department of Electronics and Communication Engineering P.S.R.Engineering CollegeSivakasi626140India

Social Networking Sites(SNSs)are nowadays utilized by the whole world to share ideas,images,and valuable contents by means of a post to reach a group of *** use of SNS often inﬂicts the physical and the mental health of the ***,researchers often focus on identifying the illegal beha-viors in the SNS to reduce its negative inﬂ*** state-of-art Natural Language processing techniques for anomaly detection have utilized a wide anno-tated corpus to identify the anomalies and they are often time-consuming as well as certainly do not guarantee maximum *** overcome these issues,the proposed methodology utilizes a Modiﬁed Convolutional Neural Network(MCNN)using stochastic pooling and a Leaky Rectiﬁed Linear Unit(LReLU).Here,each word in the social media text is analyzed based on its *** stochastic pooling accurately detects the anomalous social media posts and reduces the chance of overﬁ*** LReLU overcomes the high computational cost and gradient vanishing problem associated with other activation *** also doesn’t stop the learning process when the values are *** MCNN computes a speciﬁed score value using a novel integrated anomaly detection *** on the score value,the anomalies are identiﬁed.A Teaching Learn-ing based Optimization(TLBO)algorithm has been used to optimize the feature extraction phase of the modiﬁed CNN and fast convergence is *** this way,the performance of the model is enhanced in terms of classiﬁcation *** efﬁciency of the proposed technique is compared with the state-of-art techni-ques in terms of accuracy,sensitivity,speciﬁcity,recall,and *** proposed MCNN-TLBO technique has provided an overall architecture of 97.85%,95.45%,and 97.55%for the three social media datasets namely Facebook,Twitter,and Reddit respectively.

关键词： Anomaly detection convolutional neural network social networking sites stochastic pooling teacher learner-based optimization

来源：评论

学校读者我要写书评

暂无评论

CHIE: Generative MRC Evaluation for in-context QA with Correctness, Helpfulness, Irrelevancy, and Extraneousness Aspects 2

CHIE: Generative MRC Evaluation for in-context QA with Corre...

引用

2nd Workshop on Generalisation (Benchmarking) in NLP, GenBench 2024

作者： Phatthiyaphaibun, Wannaphong Nonesung, Surapon Limkonchotiwat, Peerat Udomcharoenchaikit, Can Sawatphol, Jitkapat Chuangsuwanich, Ekapol Nutanong, Sarana School of Information Science and Technology VISTEC Thailand Department of Computer Engineering Faculty of Engineering Chulalongkorn University Thailand

ISBN: (纸本)9798891761827

The evaluation of generative models in Machine Reading Comprehension (MRC) presents distinct difficulties, as traditional metrics like BLEU, ROUGE, METEOR, Exact Match, and F1 score often struggle to capture the nuanced and diverse responses. While embedding-based metrics such as BERTScore and BARTScore focus on semantic similarity, they still fail to fully address aspects such as recognizing additional helpful information and rewarding contextual faithfulness. Recent advances in large language model (LLM) based metrics offer more fine-grained evaluations, but challenges such as score clustering remain. This paper introduces a multi-aspect evaluation framework, CHIE, incorporating aspects of Correctness, Helpfulness, Irrelevance, and Extraneousness. Our approach, which uses binary categorical values rather than continuous rating scales, aligns well with human judgments, indicating its potential as a comprehensive and effective evaluation method. © 2024 Association for Computational Linguistics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Siamese Multiple Attention Temporal Convolution Networks for Human Mobility Signature Identification 27

Siamese Multiple Attention Temporal Convolution Networks for...

引用

27th IEEE International Conference on Intelligent Transportation Systems, ITSC 2024

作者： Zheng, Zhipeng Jiang, Yuchen Zhang, Shiyao Wei, Xuetao Research Institute for Trustworthy Autonomous Systems Southern University of Science and Technology Department of Computer Science and Engineering China

ISBN: (纸本)9798331505929

The Human Mobility Signature Identification (HuMID) problem stands as a fundamental task within the realm of driving style representation, dedicated to discerning latent driving behaviors and preferences from diverse driver trajectories for driver identification. Its solutions hold significant implications across various domains (e.g., ride-hailing, insurance), wherein their application serves to safeguard users and mitigate potential fraudulent activities. Present HuMID solutions often exhibit limitations in adaptability when confronted with lengthy trajectories, consequently incurring substantial computational overhead. Furthermore, their inability to effectively extract crucial local information further impedes their performance. To address this problem, we propose a Siamese Multiple Attention Temporal Convolutional Network (Siamese MA-TCN) to capitalize on the strengths of both TCN architecture and multi-head self-attention, enabling the proficient extraction of both local and long-term dependencies. Additionally, we devise a novel attention mechanism tailored for the efficient aggregation of multi-scale representations derived from our model. Experimental evaluations conducted on two real-world taxi trajectory datasets reveal that our proposed model effectively extracts both local key information and long-term dependencies. These findings highlight the model's outstanding generalization capabilities, demonstrating its robustness and adaptability across datasets of varying sizes. © 2024 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Navigating the Landscape of Ransomware Detection Methods: A Review 3rd

Navigating the Landscape of Ransomware Detection Methods: A ...

引用

3rd International Conference on Advanced Computing and Intelligent Technologies, ICACIT 2023

作者： Benito, Nikiema Martial, Ouedraogo Banda, Tamiano Sahana, Subrata Department of Computer Science and Engineering School of Engineering and Technology Sharda University UP Greater Noida201306 India

ISBN: (纸本)9789819719600

The global impact of ransomware on cybersecurity has increased alarmingly in recent years. It is the cause of important financial damage for individuals as well as for corporations. From the early days of computers, this type of malware has never ceased to evolve, finding new ways to evade current anti-ransomware detection. Ransomware is a malicious code or piece of code designed to extort payment from victims. Considerable advances have been done on ransomware detection. Nevertheless, many of the proposed solutions fall short because of the continuous sophistication of ransomware. This paper provides a comprehensive review of the recent evolution of ransomware and a critical analysis of state-of-art research contributions based on well-defined evaluation metrics. Our analysis covers various ransomware detection techniques and contrasts their merits and limitations. This review goal is to be a constructive guide for ransomware-oriented research. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Developing a Translation Model to Bridge ASL and Indian Regional Languages Using Neural Networks

Developing a Translation Model to Bridge ASL and Indian Regi...

引用

2025 International Conference on Computing for Sustainability and Intelligent Future, COMP-SIF 2025

作者： Swetha, M.S. Muneshwara, M.S. Venkatesh, Varnitha Kudremane, Shreyaa S. M. Nair, Ritika S Vismitha, H. Bms Institute of Technology & Management Department of Information Science and Engineering Bengaluru India Bms Institute of Technology & Management Department of Computer Science and Engineering Bengaluru India

ISBN: (纸本)9798331538538

In this resarch work, a real-time hand gesture detection system for Ame rican Sign Language (ASL) that enables the deaf-dumb co mmun ity to interact with others by translating sign language into text. The current methods for translating ASL to English primarily use computer vision tools like Media Pipe and OpenPose to recognise hand landma rks and motions. To tackle this issue, we provide a novel convolutional neural network (CNN) architecture that removes the need for manual feature engineering by automatically obtaining discriminative spatial-temporal characteristics from raw video streams. CNNs are more suited to handle the dynamic and highly changeable nature of hand movements because they are very good in simu Itaneously recording spatial and temporal informat ion, unlike LSTM (Long Short-Term Memory) networks, which are frequently emp loyed for sequential data. CNNs are perfect for rea l-t ime applications because they provide more efficient processing and faster inference time. By using multilingual language translators and advanced CNNbased deep learning models, we aim to reduce the communication gaps across India's diverse linguistic minorities. To ensure accurate performance evaluation and successful model learning, the dataset was divided into 70 % for tra inin g and 30 % for testing. Compared to tradit ional methods, the proposed approach has a 95% accuracy rate. With potential uses in public services, healthcare, and education, this program encourages assistive technology by offering a practical tool to improve communicat ion between sign language users and the general public. Through encouraging interlanguage contact, the system helps create a society that is more inclusive. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Sleep Stage and Sleep Disorder Detection using Spatial and Temporal Attention based Capsule Network 1

Sleep Stage and Sleep Disorder Detection using Spatial and T...

引用

1st International Conference on Software, Systems and Information technology, SSITCON 2024

作者： Devi, R. Renuga Elavarasi, S. Anitha Venkateswarlu, Chetti Anitha, R. Ali, Hassan Mohamed Srm Institute of Science and Technology Ramapuram Department of Computer Science and Applications Chennai India Sona College of Technology Department of Computer Science and Engineering Salem India Maisammaguda Department of Electronics and Communication Engineering Hyderabad India New Prince Shri Bhavani College of Engineering and Technology Department of Information Technology Chennai India The Islamic University College of Technical Engineering Department of Computers Techniques Engineering Najaf Iraq

ISBN: (纸本)9798350352931

Sleep is referred as a cyclic process which is the most important part in human life. The sleep disorder and its stages are observed from time waveforms such as Electroencephalogram (EEG), Electrocardiograms (ECG), and Electromyograms (EMG) while diagnosing patients. However, the existing Deep Learning (DL) based neural network used for multi-modal sleep stage and sleep disorder detection has limitations in decision making that results in inaccurate detection of sleep stage. To overcome this limitation, a Spatial and Temporal Attention based Capsule Network (STA-CapsNet) is proposed for sleep stage and sleep disorder detecting for multimodal waveforms. The proposed CapsNet model with spatial and temporal attention mechanism focuses on informative regions that helps in making decision and classifies the sleep stages accurately. Firstly, the time waveforms are acquired and preprocessed to convert the signals into Red Green Blue (RGB) images for precise classification of sleep disorders. These RGB images are fed to proposed STA-CapsNet model for classifying the six stages of sleep correctly. Experimental results of the proposed STA-CapsNet method achieved accuracy of 9 6. 8 9 %which is higher than the existing approaches such as Convolutional Neural Network (CNN) and Visual Geometry Group-16 (VGG-16). © 2024 IEEE.

关键词： Sleep research

来源：评论

学校读者我要写书评

暂无评论

Deep Neural Networks for Detecting Insider Threats and Social engineering Attacks 4

Deep Neural Networks for Detecting Insider Threats and Socia...

引用

4th IEEE International Conference on Electrical, computer, and Energy Technologies, ICECET 2024

作者： Zewdie, Marshet Girma, Anteneh Sitote, Tilahun Melak Science and Technology University Department of Computer Science & Engineering Adama Adama Ethiopia WashingtonDC United States

ISBN: (纸本)9798350395914

Cybersecurity (CS) plays a crucial role in protecting valuable and sensitive organizational data, systems, computers, and networks from unauthorized access. However, the incressing prevalence of insider threats and social engineering attack (SEA) presents significant challenges in effectively detecting and mitigating of these risks. A yearly report from 2023 highlighted that despite 90% of companies implementing multiple security measures, they still experienced an average loss of 16 million per incident. The detection capabilities of existing detection methods, which are primarily network-based or host-based intrusion detection, have limitations. This article aims to enhance detection methods through a comprehensive analysis of network and host level insiders' behavior along with Deep Learning approaches. This proposed method of detection provide a unified and holistic detection. Insider threats, whether intentional or unintentional, also create vulnerabilities to external threats and attacks such as phishing and SEA attacks. By addressing the gap in insider threat detection, the proposed comprehensive analysis of insider network and host level activities will enhance detection performance and reduce security costs by compact the existing fragmented detection approaches. As a result the false positive and false negative alarms will reduce the cost of detection and mitigate business operation disturbances. Since insiders interact with network devices and computers as users, integrating their host and network behaviors' into the detection methods offer both enhanced detection capabilities and a unified detection. To evaluate the proposed detection method, an Auto-encoder Deep Learning model will be developed, and public network and host intrusion detesets will be utilized. Evaluation metrics such as Accuracy, precision, recall, and F1- score will be employed. Preliminary analysis results have shown the proposed compre-hensive behavior analysis with Deep Learning (DL)

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Distance Matrix and Markov Chain Based Sensor Localization in WSN

引用

computers, Materials & Continua 2022年第5期71卷 4051-4068页

作者： Omaima Bamasaq Daniyal Alghazzawi Surbhi Bhatia Pankaj Dadheech Farrukh Arslan Sudhakar Sengan Syed Hamid Hassan Department of Computer Science Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddahSaudi Arabia Department of Information Systems Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddahSaudi Arabia Department of Information Systems College of Computer Science and Information TechnologyKing Faisal UniversitySaudi Arabia Department of Computer Science and Engineering Swami Keshvanand Institute of TechnologyManagement&GramothanJaipurRajasthanIndia University of Engineering and Technology LahorePakistan Department of Computer Science and Engineering PSN College of Engineering and TechnologyTirunelveliTamilnaduIndia

Applications based on Wireless Sensor Networks(WSN)have shown to be quite useful in monitoring a particular geographic area of *** geometries of the surrounding environment are essential to establish a successful WSN *** it is literally hard because constructing a localization algorithm that tracks the exact location of Sensor Nodes(SN)in a WSN is always a challenging *** this research paper,Distance Matrix and Markov Chain(DM-MC)model is presented as node localization technique in which Distance Matrix and Estimation Matrix are used to identify the position of the *** method further employs a Markov Chain Model(MCM)for energy optimization and interference *** are performed against two well-known models,and the results demonstrate that the proposed algorithm improves performance by using less network resources when compared to the existing *** probability is used in the Markova chain to sustain higher energy ***,the proposed Distance Matrix and Markov Chain model decrease energy use by 31%and 25%,respectively,compared to the existing DV-Hop and CSA *** experimental results were performed against two proven models,Distance VectorHop Algorithm(DV-HopA)and Crow Search Algorithm(CSA),showing that the proposed DM-MC model outperforms both the existing models regarding localization accuracy and Energy Consumption(EC).These results add to the credibility of the proposed DC-MC model as a better model for employing node localization while establishing a WSN framework.

关键词： Wireless sensor network resource optimization routing distance matrix Markov chain

来源：评论

学校读者我要写书评

暂无评论

Fast and Accurate Community Detection in Social Media Networks with a Multi-Algorithmic Approach

Fast and Accurate Community Detection in Social Media Networ...

引用

2024 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems, ICSES 2024

作者： Pabitha, P. Petchiammal, G. Eswari Anna University Department of Computer Technology MIT Campu Chromepet600044 India Knowledge Institute of Technology Department of Computer Science and Engineering Salem637504 India

ISBN: (纸本)9798331543617

Community detection is an essential technique for uncovering structural patterns in complex networks, such as social, biological, and information networks. In this paper to explore a hybrid approach to community detection by combining the label propagation algorithm and leiden algorithm. Although LPA is well known for its computational effectiveness in large- scale networks, because of its reliance on local label propagation, it frequently results in unstable and fractured communities. High- quality community structures are produced by the Leiden algorithm, which is more computationally expensive than the Louvain method but ensures well-connected communities by optimising modularity. Our goal in combining these two approaches is to take use of LPA's speed for quick community discovery and the Leiden algorithm's refinement powers for increased accuracy and stability. Large or dynamic networks where both real-time speed and community quality are crucial are best suited for the suggested hybrid method. The combination of these algorithms yields more cohesive and modular communities than either one working alone, as evidenced by experimental results on synthetic social network data. The method is a useful tool for large- scale network analysis because it strikes a balance between the necessity for strong and significant community structures and computational efficiency. © 2024 IEEE.

关键词： Community detection Label propagation leidens Louvain

来源：评论

学校读者我要写书评

暂无评论

RUL Estimation for Power Electronic Devices Using RNNs

RUL Estimation for Power Electronic Devices Using RNNs

引用

2024 Prognostics and System Health Management Conference, PHM 2024

作者： Cai, Congrui Lu, Zhonghai School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Embedded Systems Master Program Sweden School of Electrical Engineering and Computer Science Department of Electrical Engineering KTH Royal Institute of Technology Sweden

ISBN: (纸本)9798350360585

Recurrent Neural Networks (RNNs) are commonly used in data-driven approaches to estimate the Remaining Useful Lifetime (RUL) of power electronic devices. RNNs are preferred because their intrinsic feedback mechanisms are better suited to model time-series data. However, the impact of RNN complexity on estimation accuracy is rarely discussed in the literature. This issue is important because choosing a lower-complexity model that delivers the same or similar performance as a higher-complexity model can increase implementation efficiency. In the paper, we use three RNN models, namely, the vanilla version, LSTM (Long Short Term Memory) and GRU (Gated Recurrent Unit) to conduct RUL estimation for power electronic devices. We use two accelerated aging datasets, one dataset targeting the package failure of MOSFETs, and the other dataset targeting package failure of power diodes. Our study shows that a lower-complexity RNN does not necessarily deliver a lower performance. Similarly, a higher-complexity model does not assure a higher performance. As such, our work highlights the importance of selecting a proper neural network for RUL estimation not biased towards complex models. This is especially useful and important for implementing such RUL estimation techniques in embedded resource-constrained and speed-limited computins platforms. © 2024 IEEE.

关键词： Long short-term memory

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：