Pub Date : 2024-09-18DOI: 10.1016/j.jksuci.2024.102193
When emergencies occur, panic spreads quickly across cyberspace and physical space. Despite widespread attention to emotional contagion in cyber–physical societies (CPS), existing studies often overlook individual relationship heterogeneity, which results in imprecise models. To address this issue, we propose a heterogeneous emotional contagion method for CPS. First, we introduce the Strong–Weak Emotional Contagion Model (SW-ECM) to simulate the heterogeneous emotional contagion process in CPS. Second, we formulate the mean-field equations for the SW-ECM to accurately capture the dynamic evolution of heterogeneous emotional contagion in the CPS. Finally, we construct a small-world network based on strong–weak relationships to validate the effectiveness of our method. The experimental results show that our method can effectively simulate the heterogeneous emotional contagion and capture changes in relationships between individuals, providing valuable guidance for crowd evacuations prone to emotional contagion.
{"title":"Heterogeneous emotional contagion of the cyber–physical society","authors":"","doi":"10.1016/j.jksuci.2024.102193","DOIUrl":"10.1016/j.jksuci.2024.102193","url":null,"abstract":"<div><p>When emergencies occur, panic spreads quickly across cyberspace and physical space. Despite widespread attention to emotional contagion in cyber–physical societies (CPS), existing studies often overlook individual relationship heterogeneity, which results in imprecise models. To address this issue, we propose a heterogeneous emotional contagion method for CPS. First, we introduce the Strong–Weak Emotional Contagion Model (SW-ECM) to simulate the heterogeneous emotional contagion process in CPS. Second, we formulate the mean-field equations for the SW-ECM to accurately capture the dynamic evolution of heterogeneous emotional contagion in the CPS. Finally, we construct a small-world network based on strong–weak relationships to validate the effectiveness of our method. The experimental results show that our method can effectively simulate the heterogeneous emotional contagion and capture changes in relationships between individuals, providing valuable guidance for crowd evacuations prone to emotional contagion.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002829/pdfft?md5=f933d896a76a94be422b19df9a07b8ff&pid=1-s2.0-S1319157824002829-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142272806","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-16DOI: 10.1016/j.jksuci.2024.102191
Navigating through a tactile paved footpath surrounded by various sizes of static and dynamic obstacles is one of the biggest impediments visually impaired people face, especially in Dhaka, Bangladesh. This problem is important to address, considering the number of accidents in such densely populated footpaths. We propose a novel deep-edge solution using Computer Vision to make people aware of the obstacles in the vicinity and reduce the necessity of a walking cane. This study introduces a diverse novel tactile footpath dataset of Dhaka covering different city areas. Additionally, existing state-of-the-art deep neural networks for object detection have been fine-tuned and investigated using this dataset. A heuristic-based breadth-first navigation algorithm (HBFN) is developed to provide navigation directions that are safe and obstacle-free, which is then deployed in a smartphone application that automatically captures images of the footpath ahead to provide real-time navigation guidance delivered by speech. The findings from this study demonstrate the effectiveness of the object detection model, YOLOv8s, which outperformed other benchmark models on this dataset, achieving a high mAP of 0.974 and an F1 score of 0.934. The model’s performance is analyzed after quantization, reducing its size by 49.53% while retaining 98.97% of the original mAP.
{"title":"A novel edge intelligence-based solution for safer footpath navigation of visually impaired using computer vision","authors":"","doi":"10.1016/j.jksuci.2024.102191","DOIUrl":"10.1016/j.jksuci.2024.102191","url":null,"abstract":"<div><p>Navigating through a tactile paved footpath surrounded by various sizes of static and dynamic obstacles is one of the biggest impediments visually impaired people face, especially in Dhaka, Bangladesh. This problem is important to address, considering the number of accidents in such densely populated footpaths. We propose a novel deep-edge solution using Computer Vision to make people aware of the obstacles in the vicinity and reduce the necessity of a walking cane. This study introduces a diverse novel tactile footpath dataset of Dhaka covering different city areas. Additionally, existing state-of-the-art deep neural networks for object detection have been fine-tuned and investigated using this dataset. A heuristic-based breadth-first navigation algorithm (HBFN) is developed to provide navigation directions that are safe and obstacle-free, which is then deployed in a smartphone application that automatically captures images of the footpath ahead to provide real-time navigation guidance delivered by speech. The findings from this study demonstrate the effectiveness of the object detection model, YOLOv8s, which outperformed other benchmark models on this dataset, achieving a high mAP of 0.974 and an F1 score of 0.934. The model’s performance is analyzed after quantization, reducing its size by 49.53% while retaining 98.97% of the original mAP.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002805/pdfft?md5=67af390c0280c8b6ae2c05684fbae69f&pid=1-s2.0-S1319157824002805-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142272843","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-13DOI: 10.1016/j.jksuci.2024.102181
Incomplete knowledge graphs are common problem that can impair task accuracy. As knowledge graphs grow extensively, the probability of incompleteness increases. Link prediction addresses this problem, but accurate and efficient link prediction methods are needed to handle incomplete and extensive knowledge graphs. This study proposed modifications to the embedding-based link prediction using clustering to improve performance. The proposed method involves four main processes: embedding, clustering, determining clusters, and scoring. Embedding converts entities and relations into vectors while clustering groups these vectors. Selected clusters are determined based on the shortest distance between the centroid and the incomplete knowledge graph. Scoring measures relation rankings, and link prediction result is selected based on highest scores. The link prediction performance is evaluated using Hits@1, Mean Rank, Mean Reciprocal Rank and prediction time on three knowledge graph datasets: WN11, WN18RR, and FB13. The link prediction methods used are TransE and ComplEx, with BIRCH as the clustering technique and Mahalanobis for short-distance measurement. The proposed method significantly improves link prediction performance, achieving accuracy up to 98% and reducing prediction time by 99%. This study provides effective and efficient solution for improving link prediction, demonstrating high accuracy and efficiency in handling incomplete and extensive knowledge graphs.
{"title":"Improving embedding-based link prediction performance using clustering","authors":"","doi":"10.1016/j.jksuci.2024.102181","DOIUrl":"10.1016/j.jksuci.2024.102181","url":null,"abstract":"<div><p>Incomplete knowledge graphs are common problem that can impair task accuracy. As knowledge graphs grow extensively, the probability of incompleteness increases. Link prediction addresses this problem, but accurate and efficient link prediction methods are needed to handle incomplete and extensive knowledge graphs. This study proposed modifications to the embedding-based link prediction using clustering to improve performance. The proposed method involves four main processes: embedding, clustering, determining clusters, and scoring. Embedding converts entities and relations into vectors while clustering groups these vectors. Selected clusters are determined based on the shortest distance between the centroid and the incomplete knowledge graph. Scoring measures relation rankings, and link prediction result is selected based on highest scores. The link prediction performance is evaluated using Hits@1, Mean Rank, Mean Reciprocal Rank and prediction time on three knowledge graph datasets: WN11, WN18RR, and FB13. The link prediction methods used are TransE and ComplEx, with BIRCH as the clustering technique and Mahalanobis for short-distance measurement. The proposed method significantly improves link prediction performance, achieving accuracy up to 98% and reducing prediction time by 99%. This study provides effective and efficient solution for improving link prediction, demonstrating high accuracy and efficiency in handling incomplete and extensive knowledge graphs.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002702/pdfft?md5=e31143cd70a22f8ffa2da3a54e983856&pid=1-s2.0-S1319157824002702-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142241566","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-11DOI: 10.1016/j.jksuci.2024.102188
Recently, adversarial patches have become frequently used in adversarial attacks in real-world settings, evolving into various shapes and numbers. However, existing defense methods often exhibit limitations in addressing specific attacks, datasets, or conditions. This underscores the demand for versatile and robust defenses capable of operating across diverse scenarios. In this paper, we propose the RAPID (Robust multi-pAtch masker using channel-wise Pooled varIance with two-stage patch Detection) framework, a stable solution to restore detection efficacy in the presence of multiple patches. The RAPID framework excels in defending against attacks regardless of patch number or shape, offering a versatile defense adaptable to diverse adversarial scenarios. RAPID employs a two-stage strategy to identify and mask coordinates associated with patch attacks. In the first stage, we propose the ‘channel-wise pooled variance’ to detect candidate patch regions. In the second step, upon detecting these regions, we identify dense areas as patches and mask them accordingly. This framework easily integrates into the preprocessing stage of any object detection model due to its independent structure, requiring no modifications to the model itself. Evaluation indicates that RAPID enhances robustness by up to 60% compared to other defenses. RAPID achieves mAP50 and mAP@50-95 values of 0.696 and 0.479, respectively.
最近,对抗性补丁在现实世界的对抗性攻击中被频繁使用,并演变成各种形状和数量。然而,现有的防御方法在应对特定攻击、数据集或条件时往往表现出局限性。这凸显了对能够在不同场景下运行的多功能、强大的防御系统的需求。在本文中,我们提出了 RAPID(Robust multi-pAtch masker using channel-wise Pooled varIance with two-stage patch Detection)框架,这是一种在存在多个补丁的情况下恢复检测功效的稳定解决方案。RAPID 框架在抵御攻击方面表现出色,无论补丁数量或形状如何,都能提供适应不同对抗场景的多功能防御。RAPID 采用两阶段策略来识别和屏蔽与补丁攻击相关的坐标。在第一阶段,我们提出了 "信道汇集方差 "来检测候选补丁区域。第二步,在检测到这些区域后,我们将密集区域识别为补丁,并对其进行相应的屏蔽。由于该框架结构独立,无需修改模型本身,因此可轻松集成到任何物体检测模型的预处理阶段。评估结果表明,与其他防御方法相比,RAPID 增强了高达 60% 的鲁棒性。RAPID 的 mAP50 和 mAP@50-95 值分别为 0.696 和 0.479。
{"title":"RAPID: Robust multi-pAtch masker using channel-wise Pooled varIance with two-stage patch Detection","authors":"","doi":"10.1016/j.jksuci.2024.102188","DOIUrl":"10.1016/j.jksuci.2024.102188","url":null,"abstract":"<div><p>Recently, adversarial patches have become frequently used in adversarial attacks in real-world settings, evolving into various shapes and numbers. However, existing defense methods often exhibit limitations in addressing specific attacks, datasets, or conditions. This underscores the demand for versatile and robust defenses capable of operating across diverse scenarios. In this paper, we propose the RAPID (<strong>R</strong>obust multi-p<strong>A</strong>tch masker using channel-wise <strong>P</strong>ooled var<strong>I</strong>ance with two-stage patch <strong>D</strong>etection) framework, a stable solution to restore detection efficacy in the presence of multiple patches. The RAPID framework excels in defending against attacks regardless of patch number or shape, offering a versatile defense adaptable to diverse adversarial scenarios. RAPID employs a two-stage strategy to identify and mask coordinates associated with patch attacks. In the first stage, we propose the ‘channel-wise pooled variance’ to detect candidate patch regions. In the second step, upon detecting these regions, we identify dense areas as patches and mask them accordingly. This framework easily integrates into the preprocessing stage of any object detection model due to its independent structure, requiring no modifications to the model itself. Evaluation indicates that RAPID enhances robustness by up to 60% compared to other defenses. RAPID achieves mAP50 and mAP@50-95 values of 0.696 and 0.479, respectively.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002775/pdfft?md5=097312e661d7cf2bd4bcbc118fd164bd&pid=1-s2.0-S1319157824002775-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142169352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-11DOI: 10.1016/j.jksuci.2024.102184
Sharding is a critical technology for enhancing blockchain scalability. However, existing sharding blockchain protocols suffer from a high cross-shard ratio, high transaction latency, limited throughput enhancement, and high account migration. To address these problems, this paper proposes a sharding blockchain protocol for enhanced scalability and performance optimization through account transaction reconfiguration. Firstly, we construct a blockchain transaction account graph network structure to analyze transaction account correlations. Secondly, a modularity-based account transaction reconfiguration algorithm and a detailed account reconfiguration process is designed to minimize cross-shard transactions. Finally, we introduce a transaction processing mechanism for account transaction reconfiguration in parallel with block consensus uploading, which reduces the reconfiguration time overhead and system latency. Experimental results demonstrate substantial performance improvements compared to existing shard protocols: up to a 34.7% reduction in cross-shard transaction ratio, at least an 83.2% decrease in transaction latency, at least a 52.7% increase in throughput and a 7.8% decrease in account migration number. The proposed protocol significantly enhances the overall performance and scalability of blockchain, providing robust support for blockchain applications in various fields such as financial services, supply chain management, and industrial Internet of Things. It also enables better support for high-concurrency scenarios and large-scale network environments.
{"title":"A sharding blockchain protocol for enhanced scalability and performance optimization through account transaction reconfiguration","authors":"","doi":"10.1016/j.jksuci.2024.102184","DOIUrl":"10.1016/j.jksuci.2024.102184","url":null,"abstract":"<div><p>Sharding is a critical technology for enhancing blockchain scalability. However, existing sharding blockchain protocols suffer from a high cross-shard ratio, high transaction latency, limited throughput enhancement, and high account migration. To address these problems, this paper proposes a sharding blockchain protocol for enhanced scalability and performance optimization through account transaction reconfiguration. Firstly, we construct a blockchain transaction account graph network structure to analyze transaction account correlations. Secondly, a modularity-based account transaction reconfiguration algorithm and a detailed account reconfiguration process is designed to minimize cross-shard transactions. Finally, we introduce a transaction processing mechanism for account transaction reconfiguration in parallel with block consensus uploading, which reduces the reconfiguration time overhead and system latency. Experimental results demonstrate substantial performance improvements compared to existing shard protocols: up to a 34.7% reduction in cross-shard transaction ratio, at least an 83.2% decrease in transaction latency, at least a 52.7% increase in throughput and a 7.8% decrease in account migration number. The proposed protocol significantly enhances the overall performance and scalability of blockchain, providing robust support for blockchain applications in various fields such as financial services, supply chain management, and industrial Internet of Things. It also enables better support for high-concurrency scenarios and large-scale network environments.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002738/pdfft?md5=107fe417689144e59c75fddd0f5b671f&pid=1-s2.0-S1319157824002738-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142169351","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-10DOI: 10.1016/j.jksuci.2024.102186
Conventional multi-scroll chaotic systems are often constrained by the number of attractors and the complexity of generation, making it challenging to meet the increasing demands of communication and computation. This paper revolves around the modified Chua’s system. By modifying its differential equation and introducing traditional nonlinear functions, such as the step function sequence and sawtooth function sequence. A nested grid multi-scroll chaotic system (NGMSCS) can be established, capable of generating nested grid multi-scroll attractors. In contrast to conventional grid multi-scroll chaotic attractors, scroll-like phenomena can be initiated outside the grid structure, thereby revealing more complex dynamic behavior and topological features. Through the theoretical design and analysis of the equilibrium point of the system and its stability, the number of saddle-focused equilibrium points of index 2 is further expanded, which can generate (2 N+2) × M attractors, and the formation mechanism is elaborated and verified in detail. In addition, the generation of an arbitrary number of equilibrium points in the y-direction is achieved by transforming the x and y variables, which can generate M×(2 N+2) attractors, increasing the complexity of the system. The system’s dynamical properties are discussed in depth via time series plots, Lyapunov exponents, Poincaré cross sections, 0–1 tests, bifurcation diagrams, and attraction basins. The existence of attractors is confirmed through numerical simulations and FPGA-based hardware experiments.
传统的多辊混沌系统往往受制于吸引子的数量和生成的复杂性,因而难以满足日益增长的通信和计算需求。本文围绕修正的蔡氏系统展开论述。通过修改其微分方程并引入传统的非线性函数,如阶跃函数序列和锯齿函数序列。嵌套网格多卷混沌系统(NGMSCS)就可以建立起来,并能产生嵌套网格多卷吸引子。与传统的网格多卷积混沌吸引子相比,卷积现象可以在网格结构之外启动,从而显示出更复杂的动态行为和拓扑特征。通过对系统平衡点及其稳定性的理论设计和分析,进一步扩展了指数为 2 的鞍焦平衡点数量,可生成(2 N+2 )×M 个吸引子,并详细阐述和验证了其形成机理。此外,通过变换 x 和 y 变量,在 y 方向上生成任意数量的平衡点,可产生 M×(2 N+2) 个吸引子,增加了系统的复杂性。通过时间序列图、Lyapunov 指数、Poincaré 截面、0-1 检验、分岔图和吸引盆地,深入讨论了系统的动力学特性。吸引子的存在通过数值模拟和基于 FPGA 的硬件实验得到了证实。
{"title":"Design and FPGA implementation of nested grid multi-scroll chaotic system","authors":"","doi":"10.1016/j.jksuci.2024.102186","DOIUrl":"10.1016/j.jksuci.2024.102186","url":null,"abstract":"<div><p>Conventional multi-scroll chaotic systems are often constrained by the number of attractors and the complexity of generation, making it challenging to meet the increasing demands of communication and computation. This paper revolves around the modified Chua’s system. By modifying its differential equation and introducing traditional nonlinear functions, such as the step function sequence and sawtooth function sequence. A nested grid multi-scroll chaotic system (NGMSCS) can be established, capable of generating nested grid multi-scroll attractors. In contrast to conventional grid multi-scroll chaotic attractors, scroll-like phenomena can be initiated outside the grid structure, thereby revealing more complex dynamic behavior and topological features. Through the theoretical design and analysis of the equilibrium point of the system and its stability, the number of saddle-focused equilibrium points of index 2 is further expanded, which can generate (2 N+2) × M attractors, and the formation mechanism is elaborated and verified in detail. In addition, the generation of an arbitrary number of equilibrium points in the <em>y</em>-direction is achieved by transforming the <em>x</em> and <em>y</em> variables, which can generate M×(2 N+2) attractors, increasing the complexity of the system. The system’s dynamical properties are discussed in depth via time series plots, Lyapunov exponents, Poincaré cross sections, 0–1 tests, bifurcation diagrams, and attraction basins. The existence of attractors is confirmed through numerical simulations and FPGA-based hardware experiments.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002751/pdfft?md5=5a97268ac1950c4cb177bec835b9c871&pid=1-s2.0-S1319157824002751-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142233768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-02DOI: 10.1016/j.jksuci.2024.102182
Artificial intelligence (AI) research on video games primarily focused on the imitation of human-like behavior during the past few years. Moreover, to increase the perceived worth of amusement and gratification, there is an enormous rise in the demand for intelligent agents that can imitate human players and video game characters. However, the agents developed using the majority of current approaches are perceived as rather more mechanical, which leads to frustration, and more importantly, failure in engagement. On that account, this study proposes an imitation learning framework to generate human-like behavior for more precise and accurate reproduction. To build a computational model, two learning paradigms are explored, artificial neural networks (ANN) and adaptive neuro-fuzzy inference systems (ANFIS). This study utilized several variations of ANN, including feed-forward, recurrent, extreme learning machines, and regressions, to simulate human player behavior. Furthermore, to find the ideal ANFIS, grid partitioning, subtractive clustering, and fuzzy c-means clustering are used for training. The results demonstrate that ANFIS hybrid intelligence systems trained with subtractive clustering are overall best with an average accuracy of 95%, followed by fuzzy c-means with an average accuracy of 87%. Also, the believability of the obtained AI agents is tested using two statistical methods, i.e., the Mann–Whitney U test and the cosine similarity analysis. Both methods validate that the observed behavior has been reproduced with high accuracy.
{"title":"Towards the development of believable agents: Adopting neural architectures and adaptive neuro-fuzzy inference system via playback of human traces","authors":"","doi":"10.1016/j.jksuci.2024.102182","DOIUrl":"10.1016/j.jksuci.2024.102182","url":null,"abstract":"<div><p>Artificial intelligence (AI) research on video games primarily focused on the imitation of human-like behavior during the past few years. Moreover, to increase the perceived worth of amusement and gratification, there is an enormous rise in the demand for intelligent agents that can imitate human players and video game characters. However, the agents developed using the majority of current approaches are perceived as rather more mechanical, which leads to frustration, and more importantly, failure in engagement. On that account, this study proposes an imitation learning framework to generate human-like behavior for more precise and accurate reproduction. To build a computational model, two learning paradigms are explored, artificial neural networks (ANN) and adaptive neuro-fuzzy inference systems (ANFIS). This study utilized several variations of ANN, including feed-forward, recurrent, extreme learning machines, and regressions, to simulate human player behavior. Furthermore, to find the ideal ANFIS, grid partitioning, subtractive clustering, and fuzzy c-means clustering are used for training. The results demonstrate that ANFIS hybrid intelligence systems trained with subtractive clustering are overall best with an average accuracy of 95%, followed by fuzzy c-means with an average accuracy of 87%. Also, the believability of the obtained AI agents is tested using two statistical methods, i.e., the Mann–Whitney U test and the cosine similarity analysis. Both methods validate that the observed behavior has been reproduced with high accuracy.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002714/pdfft?md5=542b4e8449657f4dbd195276e5fb54c1&pid=1-s2.0-S1319157824002714-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142229614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-02DOI: 10.1016/j.jksuci.2024.102183
Exploring effective and efficient strategies for identifying influential nodes from social networks as seeds to promote the propagation of influence remains a crucial challenge in the field of influence maximization (IM), which has attracted significant research efforts. Deep learning-based approaches have been adopted as an alternative promising solution to the IM problem. However, a robust model that captures the associations between network information and node influence needs to be investigated, while concurrently considering the effects of the overlapped influence on training labels. To address these challenges, a GCNT model, which integrates Graph Convolutional Networks with Graph Transformers, is introduced in this paper to capture the intricate relationships among the topology of the network, node attributes, and node influence effectively. Furthermore, an innovative method called - is proposed to generate labels to alleviate the issue of overlapped influence spread. Moreover, a Mask mechanism specially tailored for the IM problem is presented along with an input embedding balancing strategy. The effectiveness of the GCNT model is demonstrated through comprehensive experiments conducted on six real-world networks, and the model shows its competitive performance in terms of both influence maximization and computational efficiency over state-of-the-art methods.
在影响力最大化(IM)领域,探索从社交网络中识别有影响力的节点作为种子以促进影响力传播的切实有效的策略仍然是一个重要挑战,吸引了大量研究人员的努力。基于深度学习的方法已被采用,作为解决 IM 问题的另一种有前途的方案。然而,需要研究一种能捕捉网络信息与节点影响力之间关联的稳健模型,同时考虑重叠影响力对训练标签的影响。为了应对这些挑战,本文引入了一个 GCNT 模型,该模型将图卷积网络与图变换器整合在一起,能有效捕捉网络拓扑、节点属性和节点影响力之间错综复杂的关系。此外,本文还提出了一种名为 "Greedy-LIE "的创新方法来生成标签,以缓解影响扩散重叠的问题。此外,还提出了专门针对 IM 问题的掩码机制以及输入嵌入平衡策略。通过在六个真实世界网络上进行的综合实验,证明了 GCNT 模型的有效性,而且该模型在影响力最大化和计算效率方面的表现都优于最先进的方法。
{"title":"GCNT: Identify influential seed set effectively in social networks by integrating graph convolutional networks with graph transformers","authors":"","doi":"10.1016/j.jksuci.2024.102183","DOIUrl":"10.1016/j.jksuci.2024.102183","url":null,"abstract":"<div><p>Exploring effective and efficient strategies for identifying influential nodes from social networks as seeds to promote the propagation of influence remains a crucial challenge in the field of influence maximization (IM), which has attracted significant research efforts. Deep learning-based approaches have been adopted as an alternative promising solution to the IM problem. However, a robust model that captures the associations between network information and node influence needs to be investigated, while concurrently considering the effects of the overlapped influence on training labels. To address these challenges, a GCNT model, which integrates Graph Convolutional Networks with Graph Transformers, is introduced in this paper to capture the intricate relationships among the topology of the network, node attributes, and node influence effectively. Furthermore, an innovative method called <span><math><mrow><mi>G</mi><mi>r</mi><mi>e</mi><mi>e</mi><mi>d</mi><mi>y</mi></mrow></math></span>-<span><math><mrow><mi>L</mi><mi>I</mi><mi>E</mi></mrow></math></span> is proposed to generate labels to alleviate the issue of overlapped influence spread. Moreover, a Mask mechanism specially tailored for the IM problem is presented along with an input embedding balancing strategy. The effectiveness of the GCNT model is demonstrated through comprehensive experiments conducted on six real-world networks, and the model shows its competitive performance in terms of both influence maximization and computational efficiency over state-of-the-art methods.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002726/pdfft?md5=fb687d0a26ab54db6f7c889e608384a1&pid=1-s2.0-S1319157824002726-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142149684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-01DOI: 10.1016/j.jksuci.2024.102145
This special issue is a collection of emerging trends and challenges in applying learning-driven approaches to data fabric architectures within the cloud-to-thing continuum. As data generation and processing increasingly occur at the edge, there is a growing need for intelligent, adaptive data management solutions that seamlessly operate across distributed environments. In this special issue, we received research contributions from various groups around the world. We chose the eight most appropriate and novel contributions to include in this special issue. These eight contributions were further categorized into three themes: Data Handling approaches, resource optimization and management, and security and attacks. Additionally, this editorial suggests future research directions that will potentially lead to groundbreaking insights, which could pave the way for a new era of learning techniques in Data Fabric and the Cloud-to-Thing Continuum.
{"title":"Learning-driven Data Fabric Trends and Challenges for cloud-to-thing continuum","authors":"","doi":"10.1016/j.jksuci.2024.102145","DOIUrl":"10.1016/j.jksuci.2024.102145","url":null,"abstract":"<div><p>This special issue is a collection of emerging trends and challenges in applying learning-driven approaches to data fabric architectures within the cloud-to-thing continuum. As data generation and processing increasingly occur at the edge, there is a growing need for intelligent, adaptive data management solutions that seamlessly operate across distributed environments. In this special issue, we received research contributions from various groups around the world. We chose the eight most appropriate and novel contributions to include in this special issue. These eight contributions were further categorized into three themes: Data Handling approaches, resource optimization and management, and security and attacks. Additionally, this editorial suggests future research directions that will potentially lead to groundbreaking insights, which could pave the way for a new era of learning techniques in Data Fabric and the Cloud-to-Thing Continuum.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002349/pdfft?md5=286285bbd5dfa0b63dd8785bf5349c2e&pid=1-s2.0-S1319157824002349-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142230508","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-08-31DOI: 10.1016/j.jksuci.2024.102177
The huge energy consumption of data centers in cloud computing leads to increased operating costs and high carbon emissions to the environment. Deep Reinforcement Learning (DRL) technology combines of deep learning and reinforcement learning, which has an obvious advantage in solving complex task scheduling problems. Deep Q Network(DQN)-based task scheduling has been employed for objective optimization. However, training the DQN algorithm may result in value overestimation, which can negatively impact the learning effectiveness. The replay buffer technique, while increasing sample utilization, does not distinguish between sample importance, resulting in limited utilization of valuable samples. This study proposes an enhanced task scheduling algorithm based on the DQN framework, which utilizes a more optimized Dueling-network architecture as well as Double DQN strategy to alleviate the overestimation bias and address the shortcomings of DQN. It also incorporates a prioritized experience replay technique to achieve importance sampling of experience data, which overcomes the problem of low utilization due to uniform sampling from replay memory. Based on these improved techniques, we developed an energy-efficient task scheduling algorithm called EETS (Energy-Efficient Task Scheduling). This algorithm automatically learns the optimal scheduling policy from historical data while interacting with the environment. Experimental results demonstrate that EETS exhibits faster convergence rates and higher rewards compared to both DQN and DDQN. In scheduling performance, EETS outperforms other baseline algorithms in key metrics, including energy consumption, average task response time, and average machine working time. Particularly, it has a significant advantage when handling large batches of tasks.
{"title":"EETS: An energy-efficient task scheduler in cloud computing based on improved DQN algorithm","authors":"","doi":"10.1016/j.jksuci.2024.102177","DOIUrl":"10.1016/j.jksuci.2024.102177","url":null,"abstract":"<div><p>The huge energy consumption of data centers in cloud computing leads to increased operating costs and high carbon emissions to the environment. Deep Reinforcement Learning (DRL) technology combines of deep learning and reinforcement learning, which has an obvious advantage in solving complex task scheduling problems. Deep Q Network(DQN)-based task scheduling has been employed for objective optimization. However, training the DQN algorithm may result in value overestimation, which can negatively impact the learning effectiveness. The replay buffer technique, while increasing sample utilization, does not distinguish between sample importance, resulting in limited utilization of valuable samples. This study proposes an enhanced task scheduling algorithm based on the DQN framework, which utilizes a more optimized Dueling-network architecture as well as Double DQN strategy to alleviate the overestimation bias and address the shortcomings of DQN. It also incorporates a prioritized experience replay technique to achieve importance sampling of experience data, which overcomes the problem of low utilization due to uniform sampling from replay memory. Based on these improved techniques, we developed an energy-efficient task scheduling algorithm called EETS (Energy-Efficient Task Scheduling). This algorithm automatically learns the optimal scheduling policy from historical data while interacting with the environment. Experimental results demonstrate that EETS exhibits faster convergence rates and higher rewards compared to both DQN and DDQN. In scheduling performance, EETS outperforms other baseline algorithms in key metrics, including energy consumption, average task response time, and average machine working time. Particularly, it has a significant advantage when handling large batches of tasks.</p></div>","PeriodicalId":48547,"journal":{"name":"Journal of King Saud University-Computer and Information Sciences","volume":null,"pages":null},"PeriodicalIF":5.2,"publicationDate":"2024-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S1319157824002660/pdfft?md5=a86e26e6d8a0d8a013697db9338917a5&pid=1-s2.0-S1319157824002660-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142149683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}