检索结果-内蒙古大学图书馆

16th International Conference on COMmunication Systems and NETworkS, COMSNETS 2024

作者： Mangalam, Kumar Subba, Basant Indian Institute of Technology Ropar Department of Computer Science Engineering India

ISBN: (纸本)9798350383119

As per the Anti-Phishing Working Group's (APWG) report, there were approximately 4.7 million phishing attacks in the year 2022. A significant portion of these phishing attacks were carried out by alluring unsuspecting users into visiting counterfeit web-pages, which are hosted using malicious uniform resource locators (URLs). This paper proposes a lightweight Bidirectional Long Short Term Memory (BiLSTM) based phishing URL detection framework for mitigating the threats posed by phishing attacks. The proposed framework initially splits the URLs into four distinct components namely, Protocol type, domain, sub-domain and top level domain(TLD) using a set of special characters as delimiters. Different feature values corresponding to these four components are then used to build a vocabulary database of the URL corpus. Thereafter, a customized FastText word embedding technique is used to learn numeric feature vector representations of the tokens(URL features) present in the vocabulary database. These learned feature vectors, along with the pre-processed instances of the URL corpus are then provided as input to train a BiLSTM based classifier model for detection of malicious phishing URLs. Experimental results on a proprietary URL corpus comprising 200,000 normal and phishing URL instances show that the proposed framework achieves high accuracy in detecting malicious phishing URLs with minimal computational overhead. © 2024 IEEE.

关键词： - Phishing attacks BiLSTM FastText word embedding PhishTank UNB dataset

来源：评论

学校读者我要写书评

暂无评论

Predictive Modeling of Performance Variability in HPC Applications 31

Predictive Modeling of Performance Variability in HPC Applic...

引用

31st IEEE International Conference on High Performance Computing, Data, and Analytics Workshops, HiPCW 2024

作者： Sahu, Pratham Malakar, Preeti Department of Computer Science and Engineering Indian Institute of Technology Kanpur India

ISBN: (纸本)9798331509118

Performance variability is common in HPC applications, due to factors such as job scheduling, resource contention, and network traffic. We propose a machine learning-based model to predict performance variability in high performance computing applications. Our model integrates classification and regression techniques to identify jobs with high variability and accurately predict their performance fluctuations. The model incorporates job-specific and system-wide factors such as network counters to enhance prediction accuracy. © 2024 IEEE

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Breast Cancer Segmentation Using UNet and Global Convolutional Networks 27th

Breast Cancer Segmentation Using UNet and Global Convolutio...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Thyagachandran, Anand Ahmed, Yeruru Asrar Department of Computer Science and Engineering Indian Institute of Technology Madras Chennai India

ISBN: (纸本)9783031782008

Breast ultrasound (BUS) imaging techniques have become efficient tools for cancer diagnosis. Convolutional neural network (CNN) based encoder-decoder architectures have been widely used for the automated segmentation of tumours in BUS images, assisting in breast cancer diagnoses. However, these models have limitations in capturing long-range dependencies. To overcome this limitation, various deep learning techniques, such as atrous convolution, attention mechanisms, and transformer encoder-based models, have been introduced to capture long-range dependencies in feature maps, improving segmentation accuracy by considering larger receptive fields and global context. As modelling techniques evolve, there is a shift towards more complex and intricate designs. This study proposes a simple yet effective model that combines UNet and Global Convolutional Network (GCN) architectures for breast lesion segmentation. By leveraging the GCN block, our model captures broader receptive fields with a simpler design strategy. We have demonstrated the efficacy of our approach through various experiments, including kernel size analysis, model component evaluation, and data preprocessing assessment. The proposed model has been evaluated using four-fold cross-validation with BUSI and Dataset-B datasets. Additionally, models trained on both datasets have been validated with a blind test dataset, where our model demonstrates better performance compared to state-of-the-art methods, achieving a 4.9% and 6.7% improvement in Intersection over Union (IoU) score, respectively. The robustness analysis and external validation experiments underscore the superior generalization performance of our model in breast lesion segmentation tasks. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

SmartAudit: Smart Contract Vulnerability Detection Using Transfer Learning 10th

SmartAudit: Smart Contract Vulnerability Detection Using Tra...

引用

10th International Symposium on Security and Privacy in Social Networks and Big Data, SocialSec 2024

作者： Jain, Ankur Tripathy, Somanath Department of Computer Science and Engineering Indian Institute of Technology Patna Patna India

ISBN: (纸本)9789819637737

Ethereum being a dominant player in decentralized finance, invites a vast number of smart contracts annually. Though smart contracts play a crucial role in developing decentralized applications that are safe, secure, and efficient, Ethereum lost a huge amount due to various attacks caused by different vulnerabilities within the smart contracts. Therefore, it is crucial to audit the reliability of smart contracts before their deployment. This paper proposes SmartAudit which identifies whether the smart contract is vulnerable. SmartAudit employs BERT (Bidirectional Encoder Representations from Transformers) to identify vulnerabilities in unseen smart contracts. It utilizes optimized opcode sequences derived from a pruned control flow graph as input. The results show that SmartAudit significantly improves accuracy, precision, recall and F1-score compared to the state-of-the-art models. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Smart contract

来源：评论

学校读者我要写书评

暂无评论

Pedestrian Trajectory Prediction with Missing Data: Datasets, Imputation, and Benchmarking 38

Pedestrian Trajectory Prediction with Missing Data: Datasets...

引用

38th Conference on Neural information Processing Systems, NeurIPS 2024

作者： Chib, Pranav Singh Singh, Pravendra Department of Computer Science and Engineering Indian Institute of Technology Roorkee India

Pedestrian trajectory prediction is crucial for several applications such as robotics and self-driving vehicles. Significant progress has been made in the past decade thanks to the availability of pedestrian trajectory datasets, which enable trajectory prediction methods to learn from pedestrians' past movements and predict future trajectories. However, these datasets and methods typically assume that the observed trajectory sequence is complete, ignoring real-world issues such as sensor failure, occlusion, and limited fields of view that can result in missing values in observed trajectories. To address this challenge, we present TrajImpute, a pedestrian trajectory prediction dataset that simulates missing coordinates in the observed trajectory, enhancing real-world applicability. TrajImpute maintains a uniform distribution of missing data within the observed trajectories. In this work, we comprehensively examine several imputation methods to reconstruct the missing coordinates and benchmark them for imputing pedestrian trajectories. Furthermore, we provide a thorough analysis of recent trajectory prediction methods and evaluate the performance of these models on the imputed trajectories. Our experimental evaluation of the imputation and trajectory prediction methods offers several valuable insights. Our dataset provides a foundational resource for future research on imputation-aware pedestrian trajectory prediction, potentially accelerating the deployment of these methods in real-world applications. Publicly accessible links to the datasets and code files are available at https://***/Pranav-chib/TrajImpute. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Machine Learning and Nature Inspired Algorithm Based Hybrid Approach for Detection of IoT Attacks 4

A Machine Learning and Nature Inspired Algorithm Based Hybri...

引用

4th International Conference on computer, Communication, Control and information technology, C3IT 2024

作者： Sikdar, Subinoy Kule, Malay Indian Institute of Engineering Science and Technology Department of Computer Science and Technology Howrah Shibpur India

ISBN: (纸本)9798350389128

Internet of Things (IoT) is revolutionizing our daily life, shifting manual systems to computerized ones. With the proliferation of IoT devices connected via the internet, ensuring security against malicious attacks is critical. So, the security of information is the greatest concern in modern days as the IoT devices are attacked by various malicious services. In this paper, we have proposed a hybrid attack detection method using machine learning and nature inspired algorithm. The main objective of this paper is to find out the best feature(s) as well as the best algorithms, which detect the attack accurately. We have used CICIoT2023 dataset to classify the malicious attacks and benign traffic. Decision Tree and One Rule algorithm has been applied on the dataset and both the algorithms reached the accuracy up to 98.22% and 98.99% respectively. An ensemble learning method was also developed using Random Forest algorithm. Furthermore, a hybrid attack detection approach has been implemented using Decision Tree and Cuckoo Search Algorithm, which can perform better than Random Forest model. The hybrid classifier has obtained the accuracy up to 99.40%, whereas the Random Forest model reached the accuracy up to 99.28%. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Bridging the Knowledge Gap: Improving BERT models for answering MCQs by using Ontology-generated synthetic MCQA Dataset 37

Bridging the Knowledge Gap: Improving BERT models for answer...

引用

37th International Florida Artificial Intelligence Research Society Conference, FLAIRS 2024

作者： Sahil, Sahil Kumar, P. Sreenivasa Department of Computer Science and Engineering Indian Institute of Technology Madras India

BERT-based models possess impressive language understanding capabilities but often lack domain-specific knowledge, limiting their performance on specialised tasks such as medical multiple-choice question answering (MCQA). In this paper, we study how biomedical ontologies, rich repositories of medical knowledge, can be harnessed to enhance BERT-based models for medical MCQA task. Our contributions include OntoMCQA-Gen, a system which leverages different biomedical ontologies to construct BioOntoMCQA, a large synthetic MCQA dataset. OntoMCQA-Gen exploits the subclass-class relationships, definitions of concepts, and also synonym relationships from the ontologies to create this dataset of MCQs automatically. We then use this synthetic dataset to fine-tune various BERT-based models to answer medical MCQs. We evaluated these fine-tuned BERT models on the challenging MedMCQA and MedQA datasets of questions from admission examinations for medical degrees in India and USA, respectively. Our evaluation study on these datasets shows that fine-tuning the BERT-based models on BioOntoMCQA results in significantly improved accuracy scores. BioBERT and PubMedBERT, pretrained on the large medical corpus, have also shown significant improvements with our technique of fine-tuning ontology-generated synthetic data. This finding highlights the effectiveness of incorporating biomedical ontologies to enhance the BERT-based model in the medical domain. Moreover, our results underscore the importance of using ontology-generated data along with model adaptation for specialised domains, contributing to a novel advancement in natural language processing. Copyright © 2024 by the authors.

关键词： Ontology

来源：评论

学校读者我要写书评

暂无评论

Robust Stability of Discrete-time Genetic Regulatory Networks with Reaction Diffusion

引用

IAENG International Journal of Applied Mathematics 2024年第9期54卷 1756-1764页

作者： Zou, Chengye Yang, Yongwei Li, Haifeng Shang, Yubao Liu, Yunong Information Science and Engineering Yanshan University Qinhuangdao066004 China School of Mathematics and Computer Science Tongling University Anhui Tongling244000 China Computer and Information Science Qinghai Institute of Technology Xining810016 China Information Science and Engineering Yanshan University Qinhuangdao066004 China

In the realm of discrete-time modeling for gene regulatory networks, significant focus has been placed on addressing the time lags inherent in the process of DNA transcription to RNA and the subsequent translation of mRNA to proteins. These temporal delays have been consistently incorporated into discrete gene regulatory network models. However, true gene regulatory networks are also subject to spatial variables, due to the uneven distribution of protein and mRNA concentrations. The integration of reaction-diffusion terms is thus essential to fully represent the impact of spatial dynamics on gene regulatory networks. In such networks, reaction-diffusion dynamics highlight the complex interactions between neighboring spatial regions, where closeness leads to mutual influences on their functional activities. Based on this conceptual groundwork, this study introduces a discrete-time gene regulatory network model that includes the mutual interconnections between spatial areas. To guarantee the model's robust stability, we have established delay-dependent stability criteria using carefully designed Lyapunov-Krasovskii functions, framed within the context of linear matrix inequalities. The robustness and effectiveness of our approach are demonstrated through a numerical example presented in this work. © (2024), (International Association of Engineers). All Rights Reserved.

关键词： Stability criteria

来源：评论

学校读者我要写书评

暂无评论

Remember This Event That Year? Assessing Temporal information and Understanding in Large Language Models

Remember This Event That Year? Assessing Temporal Informatio...

引用

2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024

作者： Beniwal, Himanshu Patel, Dishant Kowsik Nandagopan, D. Ladia, Hritik Yadav, Ankit Singh, Mayank Department of Computer Science and Engineering Indian Institute of Technology Gandhinagar India

ISBN: (纸本)9798891761681

Large Language Models (LLMs) are increasingly ubiquitous, yet their ability to retain and reason about temporal information remains limited, hindering their application in real-world scenarios where understanding the sequential nature of events is crucial. Our study experiments with 12 state-of-the-art models (ranging from 2B to 70B+ parameters) on a novel numerical-temporal dataset, TempUN, spanning from 10,000 BCE to 2100 CE, to uncover significant temporal retention and comprehension limitations. We propose six metrics to assess three learning paradigms to enhance temporal knowledge acquisition. Our findings reveal that open-source models exhibit knowledge gaps more frequently, suggesting a trade-off between limited knowledge and incorrect responses. Additionally, various fine-tuning approaches significantly improved performance, reducing incorrect outputs and impacting the identification of 'information not available' in the generations. The associated dataset and code are available at https://***/lingoiitgn/TempUN. © 2024 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Cross-lingual Editing in Multilingual Language Models 18

Cross-lingual Editing in Multilingual Language Models

引用

18th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2024 - Findings of EACL 2024

作者： Beniwal, Himanshu Kowsik Nandagopan, D. Singh, Mayank Department of Computer Science and Engineering Indian Institute of Technology Gandhinagar India

ISBN: (纸本)9798891760936

The training of large language models (LLMs) necessitates substantial data and computational resources, and updating outdated LLMs entails significant efforts and resources. While numerous model editing techniques (METs) have emerged to efficiently update model outputs without retraining, their effectiveness in multilingual LLMs, where knowledge is stored in diverse languages, remains an underexplored research area. This research paper introduces the cross-lingual model editing (XME) paradigm, wherein a fact is edited in one language, and the subsequent update propagation is observed across other languages. To investigate the XME paradigm, we conducted experiments using BLOOM, mBERT, and XLM-RoBERTa using the two writing scripts: Latin (English, French, and Spanish) and Indic (Hindi, Gujarati, and Bengali). The results reveal notable performance limitations of state-of-the-art METs under the XME setting, mainly when the languages involved belong to two distinct script families. These findings highlight the need for further research and development of XME techniques to address these challenges. For more comprehensive information, the dataset used in this research and the associated code are publicly available at the following URL. © 2024 Association for Computational Linguistics.

关键词： Computational linguistics

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：