检索结果-内蒙古大学图书馆

2025 International Conference on Advancement in data science, E-learning and Information System, ICADEIS 2025

作者： Luckianto, Marvin Addison, Renaldi Santoso, Alexander Agung Gunawan School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798331513320

Social media platforms, such as Twitter, produce significant quantities of textual data, providing important opportunities for the analysis of emotions. The identification of emotions within text is essential for applications including mental health monitoring, marketing strategies, and sentimentdriven recommendation systems. Traditional methods frequently encounter challenges associated with linguistic complexities and the ability to generalize across varied datasets. This paper proposes a novel method using Bidirectional Gated Recurrent Unit (BiGRU) networks with some regularization techniques such as dropout and batch normalization. The bidirectional architecture allows both forward and backward contextual information to be learned, thus it can express subtle emotional features effectively. Using a dataset of 89,832 labeled tweets from six different emotional classes - sadness, joy, love, anger, fear, and surprise - the model achieved an accuracy of 93.83%, precision of 93.83%, recall of 93.83%, and an F1-score of 93.83%. This demonstrates that integrating advanced neural architectures along with successful regularization techniques significantly improves the stability and effectiveness of emotion detection models. © 2025 IEEE.

关键词： Tweets

来源：评论

学校读者我要写书评

暂无评论

Multi-head attention graph convolutional network model:End-to-end entity and relation joint extraction based on multi-head attention graph convolutional network

引用

CAAI Transactions on Intelligence Technology 2023年第2期8卷 468-477页

作者： Zhihua Tao Chunping Ouyang Yongbin Liu Tonglee Chung Yixin Cao School of Computer Science University of South ChinaHengyangChina Department of Computer Science and technology Tsinghua UniversityBeijingChina School of Computer and Information Systems Singapore Management UniversitySingaporeSingapore

At present,the entity and relation joint extraction task has attracted more and more scholars'attention in the field of natural language processing(NLP).However,most of their methods rely on NLP tools to construct dependency trees to obtain sentence structure *** adjacency matrix constructed by the dependency tree can convey syntactic *** trees obtained through NLP tools are too dependent on the tools and may not be very accurate in contextual semantic *** the same time,a large amount of irrelevant information will cause *** paper presents a novel end-to-end entity and relation joint extraction based on the multihead attention graph convolutional network model(MAGCN),which does not rely on external *** generates an adjacency matrix through a multi-head attention mechanism to form an attention graph convolutional network model,uses head selection to identify multiple relations,and effectively improve the prediction result of overlapping *** authors extensively experiment and prove the method's effectiveness on three public datasets:NYT,WebNLG,and *** results show that the authors’method outperforms the state-of-the-art research results for the task of entities and relation extraction.

关键词： information retrieval natural language processing

来源：评论

学校读者我要写书评

暂无评论

Fast and scalable off-target assessment for CRISPR guide RNAs using partial matches

Fast and scalable off-target assessment for CRISPR guide RNA...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Schmitz, Carl Bradford, Jacob Salomone, Robert Perrin, Dimitri Queensland University of Technology School of Computer Science Centre for Data Science Brisbane Australia

ISBN: (纸本)9798350386226

The design of CRISPR-Cas9 guide RNAs requires the analysis of all potential target sites to determine the risk of unintentional edits for any given guide. A number of methods have been developed to evaluate this risk but the size of the target genome is a challenging factor. Many existing methods are limited by their need for an unfeasible amount of time or memory. In this paper, we introduce Crackling++, an update to our previous Crackling method, in which we have reduced processing time by imposing stricter matching criteria. We conducted a benchmark to compare our performance to other tools and show that Crackling++ is the fastest off-targeting scoring method and that it can scale to the largest genomes. © 2024 IEEE.

关键词： Genome

来源：评论

学校读者我要写书评

暂无评论

Traffic Light System With Embedded GPS (Global Positioning System) and GSM (Global System for Mobile Communications) Shield

引用

International Journal of Ambient Computing and Intelligence 2023年第1期14卷 1-13页

作者： Gaol, Ford Lumban Alam, Pramasiwo Franklyn, Muhammad Bio Angke, Kevins Matsuo, Tokuro Computer Science Department Bina Nusantara University Jakarta Indonesia Computer Science Department School of Computer Science Bina Nusantara University Jakarta Indonesia Advanced Institute of Industrial Technology Edogawa Japan

Artificial intelligence traffic controllers are being designed with the primary goal of enabling them to adapt to the most recent sensor data in order to perform ongoing optimizations on the signal timing plan for intersections in a network in order to reduce traffic congestions, the most pressing issue in traffic flow control at present. The authors are employing an intelligent traffic redirection technology to reduce traffic and road congestion. This would operate utilizing sensors to determine weight, with the result communicated to a traffic light PLC to control the detour. The result of experiments reduced the number of automobiles in a given time interval by 51%. There is improvement as well as on the average speed of automobiles that increase within the system by 49%. The authors also found a reduction of the average time a vehicle must wait in a system by 58%. Moreover, the implementation shows that the average wait times at junctions have been reduced by 34%. © 2023 IGI Global. All rights reserved.

关键词： Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

Elevating Image Steganography:A Fusion of MSB Matching and LSB Substitution for Enhanced Concealment Capabilities

引用

computers, Materials & Continua 2024年第5期79卷 2923-2943页

作者： Muhammad Zaman Ali Omer Riaz Hafiz Muhammad Hasnain Waqas Sharif Tenvir Ali Gyu Sang Choi Department of Information Technology The Islamia University of BahawalpurBahawalpur63100Pakistan Department of Computer Science The Islamia University of BahawalpurBahawalpur63100Pakistan School of Computer Science and Engineering Yeungnam UniversityGyeongbuk38541Korea

In today’s rapidly evolving landscape of communication technologies,ensuring the secure delivery of sensitive data has become an essential *** overcome these difficulties,different steganography and data encryption methods have been proposed by researchers to secure *** of the proposed steganography techniques achieve higher embedding capacities without compromising visual imperceptibility using LSB *** this work,we have an approach that utilizes a combinationofMost SignificantBit(MSB)matching andLeast Significant Bit(LSB)*** proposed algorithm divides confidential messages into pairs of bits and connects them with the MSBs of individual pixels using pair matching,enabling the storage of 6 bits in one pixel by modifying a maximum of three *** proposed technique is evaluated using embedding capacity and Peak Signal-to-Noise Ratio(PSNR)score,we compared our work with the Zakariya scheme the results showed a significant increase in data concealment *** achieved results of ourwork showthat our algorithmdemonstrates an improvement in hiding capacity from11%to 22%for different data samples while maintaining a minimumPeak Signal-to-Noise Ratio(PSNR)of 37 *** findings highlight the effectiveness and trustworthiness of the proposed algorithm in securing the communication process and maintaining visual integrity.

关键词： Steganography most significant bit(MSB) least significant bit(LSB) peak signal-to-noise ratio(PSNR)

来源：评论

学校读者我要写书评

暂无评论

Leveraging Eager Learning for Retinal Abnormality Classification with Local Binary Pattern Features

Leveraging Eager Learning for Retinal Abnormality Classifica...

引用

2025 International Conference on Advancement in data science, E-learning and Information System, ICADEIS 2025

作者： Silvanus Harianto, Eric Chandra Setiawan, Benny Yuda Prasetyo, Simeon Setyo Purwanto, Eko School of Computer Science Bina Nusantara University Computer Science Department Jakarta11480 Indonesia

ISBN: (纸本)9798331513320

This study investigates the integration of Local Binary Pattern (LBP) features with eager learning models for the classification of retinal abnormalities, focusing on diseases like Diabetic Retinopathy (DR) and Age-Related Macular Degeneration (AMD). Using the Retinal OCT Images (optical coherence tomography) dataset, we show that LBP is an effective technique for capturing the textural patterns associated with retinal conditions. Among various classifiers, the Random Forest model achieved the highest performance, with superior accuracy, recall, and F1 scores. This highlights the potential of LBP features combined with eager learning for robust and efficient classification of retinal diseases. The approach offers significant advantages for time-sensitive medical applications, supporting early predictions and real-time clinical workflows. Despite promising results, challenges such as addressing false positives and improving precision for overlapping conditions remain. Future work should explore deep learning models, hybrid feature extraction, and optimization strategies to enhance performance. Overall, this research demonstrates the potential for automated retinal disease detection systems, contributing to improved early diagnosis and better management of vision-threatening conditions. © 2025 IEEE.

关键词： Local binary pattern

来源：评论

学校读者我要写书评

暂无评论

A Comparative Analysis of User Experience and Educational Outcomes in Math Memory Games

A Comparative Analysis of User Experience and Educational Ou...

引用

2024 International Conference on computer Engineering, Network and Intelligent Multimedia, CENIM 2024

作者： Martin, Nicholas Nathanael, Raffel Jeriyanto, Doni Hermawan, Eric Savero Fung, Janshen Panggabean, Francisco Maruli School of Computer Science Bina Nusantara University Game Application & Technology Program Computer Science Department Jakarta Indonesia Master of Computer Science Binus Graduate Program Bina Nusantara University Jakarta Indonesia

ISBN: (纸本)9798350368802

This research explores multiple varieties of effective and child-friendly learning approaches for mathematics and memorization among young children in today's increasingly digital age considering their specific developmental needs. This type of approach includes in-person teaching, engaging online videos, interactive textbooks, and innovative educational games. This research employs a detailed comparative analysis, considering both 'User Experience' and measurable learning outcome factors, and spans a range of age groups from preschoolers to elementary school children. Through highlighting the strengths and weaknesses of each method, this research aims to inspire the adoption of efficient, child-centric, and novel teaching techniques by educators and institutions, thereby advancing the quality of education in Indonesia. The results show that using serious game in education presents high usefulness score alongside positive user experience, suggesting its viability for educational purposes. © 2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Multi-Dimensional Training Optimization for Efficient Federated Synergy Learning

引用

IEEE Transactions on Mobile Computing 2025年第7期24卷 6243-6258页

作者： Fu, Shucun Dong, Fang Chen, Runze Shen, Dian Zhang, Jinghui He, Qiang Southeast University School of Computer Science and Engineering Nanjing211189 China Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Wuhan430074 China

Edge learning (EL) is an end-to-edge collaborative learning paradigm enabling devices to participate in model training and data analysis, opening countless opportunities for edge intelligence. As a promising EL framework, federated synergy learning (FSyL) mitigates the computation and communication overhead on resource-constrained devices by offloading partial model layers to the edge server for synergistic training. Nevertheless, due to the system and statistical heterogeneity, naively using existing FSyL methods is significantly time-consuming and causes accuracy degradation. Motivated by this issue, this paper introduces a novel FSyL framework that integrates multi-dimensional training optimization and formulates the edge learning cost minimization (ELCM) problem. To tackle the ELCM efficiently, we design OL-MG, an OnLine Model Splitting and Resource Provisioning Game. Specifically, we first reformulate and decompose the original ELCM based on data quality evaluation. Then, given a model splitting decision, we determine the optimal resource provisioning in Sub-problem1, based on which optimal model splitting in Sub-problem2 is modeled as a potential game. Subsequently, we introduce a decentralized algorithm to find a Nash equilibrium (NE) solution. Furthermore, we further extend OL-MG to support a budget-aware multi-edge scenario. Extensive experiments demonstrate that the proposed mechanism significantly outperforms state-of-the-art methods in cost-saving and accuracy improvement. © 2002-2012 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Provably Secure Efficient Key-Exchange Protocol for Intelligent Supply Line Surveillance in Smart Grids

引用

IEEE Internet of Things Journal 2025年第12期12卷 21784-21792页

作者： Ayub, Muhammad Faizan Li, Xiong Mahmood, Khalid Alenazi, Mohammed J. F. Das, Ashok Kumar Wang, Guijuan University of Electronic Science and Technology of China School of Computer Science and Engineering Chengdu611731 China National Yunlin University of Science and Technology Graduate School of Intelligent Data Science Douliu64002 Taiwan King Saud University College of Computer and Information Sciences Department of Computer Engineering Riyadh12311 Saudi Arabia International Institute of Information Technology Hyderabad Center for Security Theory and Algorithmic Research Hyderabad500032 India Korea University College of Informatics Department of Computer Science and Engineering Seoul02841 Korea Republic of Key Laboratory of Computing Power Network and Information Security Shandong Provincial Key Laboratory of Industrial Network and Information System Security Shandong Fundamental Research Center for Computer Science Ministry of Education Shandong Computer Science Center Jinan250353 China

Intelligent supply line surveillance is critical for modern smart grids (SGs). Smart sensors and gateway nodes are strategically deployed along supply lines to achieve intelligent surveillance. They collect data continuously and transmit it to the control centre in real-time. It enables real-time monitoring, fault detection, and efficient energy management across distribution networks. This advanced surveillance system ensures continuous monitoring of supply lines, detecting anomalies, and optimizing operations to maintain the stability and reliability of the SG. However, the reliance of all participating nodes on public communication channels to transmit supply line surveillance data exposes these systems to critical cyber attacks. These cyber attacks include impersonation, physical tampering, ephemeral secret leakage (ESL), and desynchronization attacks. To address these issues, existing key-exchange protocols often fail to ensure robust security while imposing high computation and communication overheads, limiting their practicality for resource-constrained environments. Therefore, we propose a secure and efficient key exchange and tamper-resistant authentication protocol using elliptic curve cryptography (ECC) and physical unclonable functions (PUFs). The PUF mechanism provides robust resistance against physical tampering and cloning attacks, ensuring enhanced physical security for supply line devices. We validate the security robustness of the devised protocol using the random or real (ROR) model. Furthermore, informal security analysis demonstrates the protocol’s robustness in rigorously resisting various attacks, including physical tampering, impersonation, ESL and desynchronization. Moreover, we determine the performance evaluation that reveals the protocol’s superior efficiency compared to competing protocols, achieving significant reductions of 28.64% in computation overhead and 9.96% in communication overhead. © 2014 IEEE.

关键词： Authentication Protocol

来源：评论

学校读者我要写书评

暂无评论

Solving Sparse Reward Tasks Using Self-Balancing Exploration and Exploitation

引用

Journal of Internet Technology 2025年第3期26卷 293-301页

作者： Kong, Yan Wei, Junfeng Hsia, Chih-Hsien School of Computer and Software Nanjing University of Information Science and Technology China Department of Computer Science and Information Engineering National Ilan University Taiwan Department of Business Administration Chaoyang University of Technology Taiwan

A core challenge in applying deep reinforcement learning (DRL) to real-world tasks is the sparse reward problem, and shaping reward has been one effective method to solve it. However, due to the enormous state space and sparse rewards in the real world, a large number of useless samples may be generated, leading to reduced sample efficiency and potential local optima. To address this issue, this study proposes a self-balancing method of exploration and development to solve the issue of sparse rewards. Firstly, we shape the reward function according to the evaluated progress, to guide the agent’s learning of high-reward samples. Secondly, we construct a dual-trajectory exploration network, which provides intrinsic rewards based on the novelty of states and the trajectory difference of sibling agents to encourage the agent to explore and adjust the balance between exploration and exploitation. This method effectively prevents the generation of a large amount of useless training data during the interaction between the agent and the environment, resolves local optimal dilemmas through state novelty, and adjusts the strategy in a timely manner to solve sparse reward tasks. Our method outperforms basic reinforcement learning (RL) and curiosity-driven incentives in these experimental tasks. The self-balancing exploration and exploitation approach in our research provides a new perspective and effective solution for addressing the problem of sparse rewards, thereby advancing the application of DRL in real-world problems and achieving greater success. © 2025 Taiwan Academic Network Management Committee. All rights reserved.

关键词： Deep learning

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

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

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：