检索结果-内蒙古大学图书馆

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Xu, Zhenjie Chen, Wenqing Tang, Yi Li, Xuanying Hu, Cheng Chu, Zhixuan Ren, Kui Zheng, Zibin Lu, Zhichao School of Software Engineering Sun Yat-sen University China School of Physics and Astronomy Sun Yat-sen University China School of Cyber Science and Technology Zhejiang University China Department of Computer Science City University of Hong Kong Hong Kong

ISBN: (纸本)157735897X

Natural language processing (NLP) has seen remarkable advancements with the development of large language models (LLMs). Despite these advancements, LLMs often produce socially biased outputs. Recent studies have mainly addressed this problem by prompting LLMs to behave ethically, but this approach results in unacceptable performance degradation. In this paper, we propose a multi-objective approach within a multi-agent framework (MOMA) to mitigate social bias in LLMs without significantly compromising their performance. The key idea of MOMA involves deploying multiple agents to perform causal interventions on bias-related contents of the input questions, breaking the shortcut connection between these contents and the corresponding answers. Unlike traditional debiasing techniques leading to performance degradation, MOMA substantially reduces bias while maintaining accuracy in downstream tasks. Our experiments conducted in two datasets and two models demonstrate that MOMA reduces bias scores by up to 87.7%, with only a marginal performance degradation of up to 6.8% in the BBQ dataset. Additionally, it significantly enhances the multi-objective metric icat in the StereoSet dataset by up to 58.1%. © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Hybrid Models for Recognizing Indonesian Textual Entailment 7

Hybrid Models for Recognizing Indonesian Textual Entailment

引用

7th International Conference on Informatics and Computational Sciences, ICICoS 2024

作者： Ahmadian, Hendri Abidin, Taufik Fuadi Riza, Hammam Muchtar, Kahlil Universitas Syiah Kuala Graduate School of Mathematics and Applied Sciences Banda Aceh Indonesia Universitas Syiah Kuala Faculty of Mathematics and Natural Sciences Department of Informatics Banda Aceh Indonesia National Research and Innovation Agency Research Center for Artificial Intelligence and Cyber Security Jakarta Indonesia Universitas Syiah Kuala Faculty of Engineering Department of Electrical and Computer Engineering Banda Aceh Indonesia

ISBN: (纸本)9798350375886

A paragraph raises pertinent issues about linguistic disparities in understanding the semantic relationships between sentences within the natural language processing domain (NLP) Addressing this challenge necessitates the development of robust NLP models, with a specific emphasis on text entailment. This research endeavors to enhance the performance of Indonesian NLP benchmarks in text entailment by harnessing a variant of the Bidirectional Encoder Representations from Transformer (BERT) model tailored explicitly for the Indonesian language, known as IndoBERT. The approach in this study adopts a hybrid methodology, amalgamating the IndoBERT layer modification model by summing the last four hidden layers with neural networks. The performance evaluation of the hybrid models, combining the IndoBERT layer modification model with the Bidirectional Gated Recurrent Unit (BiGRU) model and the IndoBERT layer modification model with the Bidirectional Long Short-Term Memory (BiLSTM) model, yields the F1-scores of 0.87 and 0.84, respectively. © 2024 IEEE.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

PriRoAgg: Achieving Robust Model Aggregation With Minimum Privacy Leakage for Federated Learning

引用

IEEE Transactions on Information Forensics and security 2025年 20卷 5690-5704页

作者： Hou, Sizai Li, Songze Jahani-Nezhad, Tayyebeh Caire, Giuseppe Hong Kong University of Science and Technology Division of Emerging Interdisciplinary Areas Hong Kong Southeast University School of Cyber Science and Engineering Nanjing 211189 China Supervision and Management (Southeast University) Ministry of Education Engineering Research Center of Blockchain Application Nanjing 211189 China Technische Universität Berlin Department of Electrical Engineering and Computer Science Berlin 10587 Germany

Federated learning (FL), as a promising machine learning paradigm for large-scale distributed data, faces two security challenges of privacy and robustness: the transmitted model updates potentially leak sensitive user information, and the lack of central control over local model updates leaves the global model susceptible to malicious attacks. Current solutions attempting to address both problems under the one-server FL setting fall short in the following aspects: 1) design for simple validity checks that are insufficient against advanced attacks (e.g., checking norm of individual update);and 2) have partial privacy leakage for more complicated robust aggregation algorithms (e.g., distances between model updates are leaked for multi-Krum). In this work, we formalize a novel security notion of aggregated privacy that characterizes the minimum amount of user information, in the form of aggregated statistics of users’ updates, that is necessary to be revealed to accomplish more advanced robust aggregation. We develop a general framework PriRoAgg, utilizing Lagrange coded computing and distributed zero-knowledge proof, to execute a wide range of robust aggregation algorithms while satisfying aggregated privacy. As concrete instantiations of PriRoAgg, we construct two secure and robust protocols based on state-of-the-art robust algorithms, for which we provide full theoretical analyses on security and complexity. Extensive experiments are conducted for these protocols, demonstrating their robustness against various model integrity attacks, and their efficiency advantages over baselines. © 2005-2012 IEEE.

关键词： Byzantine resilience Federated learning privacy preservation secure and robust aggregation

来源：评论

学校读者我要写书评

暂无评论

Prediction of Hypertension Disease Using Machine Learning Techniques: Case Study from Jordan 25

Prediction of Hypertension Disease Using Machine Learning Te...

引用

25th International Arab Conference on Information technology, ACIT 2024

作者： Alazaidah, Raed Samara, Ghassan Katrawi, Anwar Hadi, Wael Al-Safarini, Maram Y. Al-Mamoori, Farah Elrashidi, Ali Alqaraleh, Muhyeeddin Alzoubi, Haneen Zarqa University Faculty of Information Technology Department of Data Science and Ai Zarqa Jordan Zarqa University Faculty of It Department of Computer Science Zarqa Jordan Petra University Faculty of It Information Security Department Amman Jordan Zarqa University Faculty of Pharmacy Department of Pharmaceutical Zarqa Jordan University of Business and Technology Department of Electrical Engineering Jeddah21432 Saudi Arabia Zarqa University Faculty of It Department of Software Engineering Zarqa Jordan Applied Science Private University Facultty of Information Technology Department of Computer Science Amman Jordan

ISBN: (纸本)9798331540012

Hypertension is one of the most common diseases in Jordan. It is one of the main reasons of death among Jordanian adult citizens. Worldwide, nearly 13% of all deaths are due to Hypertension with nearly 8 million deaths yearly. Hence, this research is interested in the early prediction of this disease among Jordanian people. To achieve this main objective, Machine Learning (ML) is utilized through a large number of classification models and considering five well-known evaluation metrics. These classification models have been trained on a primary dataset that has been collected for the purpose of the research. The results revealed that Random Forest, Random Committee, and MultilayerPerceptron are the best classification models to handle the task of the early prediction of Hypertension in Jordan. © 2024 IEEE.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Decentralized Charity Network Architecture with Hyperledger Fabric Blockchain

Decentralized Charity Network Architecture with Hyperledger ...

引用

International Conference on computer and Information technology (ICCIT)

作者： Tahia Hoque Mahmuda Chowdhury Nijhum Abdullahil Kafi Mehedi Hasan Siyam Department of Cyber Security Bangladesh University of Professionals(BUP) Dhaka Bangladesh Department of Computer Science and Engineering International Islamic University Chitttagong(IIUC) Bangladesh Ahsanullah University of Science and Technology (AUST) Dhaka Bangladesh

ISBN: (数字)9798331519094

ISBN: (纸本)9798331519100

With the use of Hyperledger Fabric framework and Blockchain technology, this research offers a trustworthy and safe charity platform that can guard against fraud and promote transparency in the nonprofit industry. Here, a permissioned Blockchain network is used to create a network that let members and authorized companies to connect via a variety of channels, each with its own certificate authority (CA). By limiting access to sensitive information to only reliable network entities, this design improves data privacy. The study improves the charity platform by increasing its effectiveness and accountability. It enables the tracking and observation of all actions, including those of contributors, beneficiaries, and charity. We have developed a trustworthy infrastructure of charitable foundations for those who donate to charities and those who get benefits from them. This will stimulate donors to donate, and its execution will facilitate the growth of robust and productive connections between contributors, beneficiaries, and charity organizations. We built an architecture where donations are tracked and funds requests are received. There are three organizations involved: auditorOrg, donorOrg, and beneficiaryOrg. Transactions take place in the channel, which is exclusively visible to these entities. The architecture operates on the blockchain, symbolizing the âBlockchainâ component, which guarantees transparency, immutability, and decentralized control over charitable transactions.

关键词： Systematics Distributed ledger Smart contracts computer architecture Reliability engineering Fabrics Real-time systems Blockchains Recording Resource management

来源：评论

学校读者我要写书评

暂无评论

Fake Currency Detection using Ensemble Learning 10

Fake Currency Detection using Ensemble Learning

引用

10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and computer engineering, UPCON 2023

作者： Kumar, Ashok Sachi, Savya Bhatia, Anshoo Belwal, Pravesh Kumar, Santosh Bhatnagar, Vivek Koneru Lakshmaiah Education Foundation Department of Computer Science & Engineering AP Vaddeswaram India Lalit Narayan Mishra College of Business Management Department of Information Technology Bihar Muzaffarpur842001 India Coer University Department of Cyber Security U.K. Roorkee India Dev Bhoomi Uttarakhand University Department of Electronics & Communication U.K. Dehradun India Noida Institute of Engineering and Technology CSE-IOT Greater Noida India Haryana Ambala India

ISBN: (纸本)9798350382471

Fake currency is one of the most serious issues that may arise during financial transactions. For a developing nation like India., this is becoming a significant obstacle. Because of advancements in printing and scanning technology., it is now feasible for anybody to create counterfeit banknotes with the use of the most up-to-date hardware equipment. Identifying false notes manually is a procedure that is both time-consuming and messy;hence., there is a need for automated approaches that make fake money detection a more effective and efficient process. For this machine learning model i.e. ensemble learning is proposed in this research for detecting the forgery currency. A data set consisting of forgery and original currency images have been chosen for the analysis. The proposed model is compared in accordance with performance parameters like accuracy., precision., recall., and f1-score. With other existing models like naive bayes., random forest., and decision tree to prove efficiency of the proposed algorithm. The evaluation of these algorithms prove that the proposed algorithm ensemble learning is highly efficient with 99% accuracy in detecting forgery currency. This can be used in some of the automated machines. © 2023 IEEE.

关键词： deep learning Ensemble learning fake currency Machine learning matlab

来源：评论

学校读者我要写书评

暂无评论

Real-Time Monitoring of Structural Health with Wireless Sensor Networks in IoT Infrastructure 3

Real-Time Monitoring of Structural Health with Wireless Sens...

引用

3rd IEEE International Conference on Automation and Computation, AUTOCOM 2025

作者： Bhushan, Bharat Pushparajesh, V. Bethanney Janney, J. Sahoo, Girija Shankar Aggarwal, Prateek Das, Dibyasundar Chitkara University Chitkara Centre for Research and Development Himachal Pradesh Baddi174103 India Karnataka Bangalore India Sathyabama Institute of Science and Technology Department of Biomedical Engineering Tamil Nadu Chennai India Maharishi School of Engineering & Technology Maharishi University of Information Technology Uttar Pradesh India Chitkara University Centre of Research Impact and Outcome Punjab Rajpura140417 India Centre for Cyber Security Odisha Bhubaneswar India

ISBN: (纸本)9798331542375

Wireless Sensor Networks for structural health monitoring of infrastructure are now becoming an IoT. It provides the ability to collect and analyze survivability data in real-time to enable continuous structural health monitoring for buildings, bridges, and other critical infrastructure. The purpose of this abstract is to introduce Real-time Monitoring with Wireless Sensor Networks in IoT Infrastructure. Wireless Sensor Networks are used in structural health monitoring. They are typically composed of many distributed, cheap sensors embedded in the structure that communicate wirelessly to a central monitoring station. Those sensors, including temperature, strain and vibration, are critical in detecting potential structural problems. This data gets transmitted to a cloud for processing, which provides indications of an unusual characteristic or predictive failure. Compared to the traditional methods, there are several advantages of applying WSN for real-time structural health monitoring. Lead leads provide real-time updates on the condition of a bridge that allows for continuous monitoring, with advanced sensors pinpointed through its structure. It will assist in keeping the infrastructure safe and affordable by guaranteeing rapid time-to-resolution. In addition, WSNs have low implementation costs and do not interfere too much with the structure of large structures. © 2025 IEEE.

关键词： Wireless sensor networks

来源：评论

学校读者我要写书评

暂无评论

Real-time behavior recognition of animal: an IoT-based system design using acceleration data

引用

Multimedia Tools and Applications 2024年 1-26页

作者： Tran, Duc-Nghia Manh, Do Viet Van Thanh, Pham Shankar, A. Achyut Joshi, Kireet Tran, Duc-Tan Institute of Information Technology Vietnam Academy of Science and Technology Hanoi Viet Nam University of Fire Prevention and Fighting Hanoi100000 Viet Nam Department of Cyber Systems Engineering WMG University of Warwick CoventryCV74AL United Kingdom Department of CSE University Centre for Research & ampDevelopment Chandigarh University Punjab Mohali140413 India School of Computer Science Engineering Lovely Professional University Punjab Phagwara144411 India Department of Computer Science and Engineering Graphic Era Deemed to be University Dehradun248002 India Center of Research Impact and Outcome Chitkara University Punjab Chandigarh India Department of Computer Science & amp Engineering Graphic Era Deemed to be University Uttarakhand Dehradun248002 India Faculty of Electrical and Electronic Engineering Phenikaa University Hanoi12116 Viet Nam

By analyzing large amounts of animal behavior data, humans can assess cattle’s condition. It leads to research and development a field using accelerometers and machine learning algorithms to ‘study’ behavior from acceleration data. It is challenging to analyze animal behavior due to the activity complexity. For example, acceleration data of some cow behaviors (e.g., feeding and standing) are similar if these behaviors contain similar gestures. In practical, real-time applications even bring more challenging since it is necessary to minimize a data window for behavior consideration. Using short data window often reduces the performance of classification algorithms. This study proposes a new solution for the mentioned challenges by combining neck-mounted and leg-mounted accelerometers, then using the vector of the dynamic body acceleration (VeDBA) and Mean features of synchronized acceleration data for the Decision Tree algorithm. The proposed method is demonstrated that performs well (0.94 accuracy, 0.91 sensitivity, 0.90 positive predictive value) within a low computational complexity algorithm and short data window (3-second data window, at the rate of one sample per second). It is particularly suited for the case of a real-time IoT-based system. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Accelerometers

来源：评论

学校读者我要写书评

暂无评论

Enhancing Pre-Trained Language Representations Based on Contrastive Learning for Unsupervised Keyphrase Extraction 34

Enhancing Pre-Trained Language Representations Based on Cont...

引用

34th International Conference on Software engineering and Knowledge engineering, SEKE 2022

作者： Wang, Zhaohui Zhang, Xinghua Li, Yanzeng Wang, Yubin Sheng, Jiawei Liu, Tingwen Xu, Hongbo Institute of Information Engineering Chinese Academy of Sciences Beijing China School of Cyber Security University of Chinese Academy of Sciences Beijing China Wangxuan Institute of Computer Technology Peking University Beijing China

ISBN: (纸本)1891706543

Keyphrase extraction (KPE) aims to obtain a set of phrases from a document that can summarize the main content of the document. Recently, pre-trained language models (LMs), especially BERT and ELMo, have achieved remarkable success, presenting new state-of-the-art results in unsupervised KPE. However, current pre-trained LMs focus on building language modeling objectives to learn a general representation, ignoring the keyphrase-related knowledge. Intuitively, the joint embedding of the keyphrase set should tend to be close to that of the extracted document, and far from those of other documents. In this work, we propose a contrastive learning-based semantic representation task to further improve BERT for unsupervised KPE. Particularly, we design a doc-phrase attention module to generate joint semantic embedding of the keyphrase set as a positive sample and select other semantically similar documents as hard negative samples. In the prediction layer, we further add an accumulated self-attention module to calculate the final scores of candidate phrases. We compare with eight strong baselines, and evaluate our model on three publicly available datasets. Experimental results show that our model is effective and robust on both long and short documents. © 2022 Knowledge Systems Institute Graduate School. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Federated Learning-based Big Data Analytics For The Education System

Federated Learning-based Big Data Analytics For The Educatio...

引用

Intelligent Computing and Emerging Communication Technologies (ICEC), International Conference on

作者： Praneetha Surapaneni Sriramulu Bojjagani Neeraj Kumar Sharma Department of Computer Science School of Engineering and Applied Sciences (SEAS) Cyber Security and IoT Lab SRM University-AP Amaravati Andhra Pradesh India

ISBN: (数字)9798331508432

ISBN: (纸本)9798331508449

This paper proposes a novel approach to enhancing education systems by integrating federated learning techniques with big data analytics. Traditional data analysis methods in educational settings often need help regarding data privacy, security, and scalability. Federated learning addresses these issues by enabling collaborative model training across distributed datasets without data centralization, thus preserving the privacy of sensitive information. By harnessing the vast amounts of educational data generated from various sources such as online learning platforms, student information systems, and academic applications, federated learning empowers educational institutions to derive valuable insights while respecting data privacy regulations. Leveraging the collective intelligence of decentralized data sources, federated learning algorithms facilitate the development of robust predictive models for student performance, personalized learning recommendations, and early intervention strategies. Moreover, federated learning enables continuous model improvement by aggregating local model updates from participating institutions, ensuring adaptability to evolving educational landscapes. This paper explores the technical foundations of federated learning, its application in education systems, and its potential benefits in improving learning outcomes and fostering data-driven decision-making in education. Through a comprehensive review of existing literature and case studies, this research aims to provide insights into the opportunities and challenges associated with implementing federated learning-based big data analytics in education systems, ultimately paving the way for a more efficient and personalized approach to education.

关键词： Training Measurement Data privacy Adaptation models Accuracy Federated learning Predictive models Big Data Data models Stakeholders

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：