Information Processing & Management最新文献_第5页

Membership inference attacks via spatial projection-based relative information loss in MLaaS 通过 MLaaS 中基于空间投影的相对信息损失进行成员推理攻击

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-11-04 DOI: 10.1016/j.ipm.2024.103947

Zehua Ding , Youliang Tian , Guorong Wang , Jinbo Xiong , Jinchuan Tang , Jianfeng Ma

Machine Learning as a Service (MLaaS) has significantly advanced data-driven decision-making and the development of intelligent applications. However, the privacy risks posed by membership inference attacks (MIAs) remain a critical concern. MIAs are primarily classified into score-based and perturbation-based attacks. The former relies on shadow data and models, which are difficult to obtain in practical applications, while the latter depends solely on perturbation distance, resulting in insufficient identification performance. To this end, we propose a Spatial Projection-based Relative Information Loss (SPRIL) MIA to ascertain the sample membership by flexibly controlling the size of perturbations in the noise space and integrating relative information loss. Firstly, we analyze the alterations in predicted probability distributions induced by adversarial perturbations and leverage these changes as pivotal features for membership identification. Secondly, we introduce a spatial projection technique that flexibly modulates the perturbation amplitude to accentuate the difference in probability distributions between member and non-member data. Thirdly, this quantifies the distribution difference by calculating relative information loss based on KL divergence to identify membership. SPRIL provides a solid method to assess the potential risks of DNN models in MLaaS and demonstrates its efficacy and precision in black-box and white-box settings. Finally, experimental results demonstrate the effectiveness of SPRIL across various datasets and model architectures. Notably, on the CIFAR-100 dataset, SPRIL achieves the highest attack accuracy and AUC, reaching 99.27% and 99.73%, respectively.

机器学习即服务（MLaaS）极大地推动了数据驱动决策和智能应用的开发。然而，成员推理攻击（MIAs）带来的隐私风险仍是一个重要问题。成员推理攻击主要分为基于分数的攻击和基于扰动的攻击。前者依赖于影子数据和模型，在实际应用中很难获得；后者则完全依赖于扰动距离，导致识别性能不足。为此，我们提出了一种基于空间投影的相对信息损失（SPRIL）MIA，通过灵活控制噪声空间中扰动的大小和整合相对信息损失来确定样本的成员资格。首先，我们分析了对抗性扰动引起的预测概率分布的变化，并利用这些变化作为成员身份识别的关键特征。其次，我们引入了一种空间投影技术，可灵活调节扰动幅度，以突出成员数据和非成员数据之间概率分布的差异。第三，通过计算基于 KL 发散的相对信息损失来量化分布差异，从而识别成员身份。SPRIL 提供了一种可靠的方法来评估 DNN 模型在 MLaaS 中的潜在风险，并证明了其在黑盒和白盒设置中的有效性和精确性。最后，实验结果证明了 SPRIL 在各种数据集和模型架构中的有效性。值得注意的是，在 CIFAR-100 数据集上，SPRIL 实现了最高的攻击准确率和 AUC，分别达到 99.27% 和 99.73%。

{"title":"Membership inference attacks via spatial projection-based relative information loss in MLaaS","authors":"Zehua Ding , Youliang Tian , Guorong Wang , Jinbo Xiong , Jinchuan Tang , Jianfeng Ma","doi":"10.1016/j.ipm.2024.103947","DOIUrl":"10.1016/j.ipm.2024.103947","url":null,"abstract":"<div><div>Machine Learning as a Service (MLaaS) has significantly advanced data-driven decision-making and the development of intelligent applications. However, the privacy risks posed by membership inference attacks (MIAs) remain a critical concern. MIAs are primarily classified into score-based and perturbation-based attacks. The former relies on shadow data and models, which are difficult to obtain in practical applications, while the latter depends solely on perturbation distance, resulting in insufficient identification performance. To this end, we propose a Spatial Projection-based Relative Information Loss (SPRIL) MIA to ascertain the sample membership by flexibly controlling the size of perturbations in the noise space and integrating relative information loss. Firstly, we analyze the alterations in predicted probability distributions induced by adversarial perturbations and leverage these changes as pivotal features for membership identification. Secondly, we introduce a spatial projection technique that flexibly modulates the perturbation amplitude to accentuate the difference in probability distributions between member and non-member data. Thirdly, this quantifies the distribution difference by calculating relative information loss based on KL divergence to identify membership. SPRIL provides a solid method to assess the potential risks of DNN models in MLaaS and demonstrates its efficacy and precision in black-box and white-box settings. Finally, experimental results demonstrate the effectiveness of SPRIL across various datasets and model architectures. Notably, on the CIFAR-100 dataset, SPRIL achieves the highest attack accuracy and AUC, reaching 99.27% and 99.73%, respectively.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103947"},"PeriodicalIF":7.4,"publicationDate":"2024-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142577975","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Higher-order structure based node importance evaluation in directed networks 基于高阶结构的有向网络节点重要性评估

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-11-04 DOI: 10.1016/j.ipm.2024.103948

Meng Li , Zhigang Wang , An Zeng , Zengru Di

Evaluating the significance of objects with possible relevant information is a crucial topic in information science. Due to the fact that objects related to each other can often be described using complex networks, this topic also forms a fundamental theme in network science. Most traditional methods for characterizing the importance of nodes in complex networks only utilize the binary relationships between node pairs, neglecting the influence brought by higher-order structures. Considering the specific interaction modes between local nodes in the network, this paper associates the higher-order structural characteristics of the network with the importance of the nodes. It constructs an evaluation framework for the importance of nodes in directed networks based on higher-order structures. Experimental analysis on both artificial data and scientific citation data from the APS dataset has validated the effectiveness of the proposed algorithms. Compared with PageRank and eigenvector centrality, the proposed algorithms demonstrated higher accuracy, revealing the role of higher-order structures in node importance evaluation. Finally, a robustness analysis of several algorithms indicated that the proposed algorithms exhibited good robustness.

评估具有可能相关信息的对象的重要性是信息科学的一个重要课题。由于相互关联的对象通常可以用复杂网络来描述，因此这一课题也构成了网络科学的一个基本主题。表征复杂网络中节点重要性的传统方法大多只利用节点对之间的二元关系，忽略了高阶结构带来的影响。考虑到网络中局部节点之间的特定交互模式，本文将网络的高阶结构特征与节点的重要性联系起来。本文构建了一个基于高阶结构的有向网络节点重要性评估框架。对人工数据和来自 APS 数据集的科学引文数据的实验分析验证了所提算法的有效性。与 PageRank 和特征向量中心性相比，所提出的算法具有更高的准确性，揭示了高阶结构在节点重要性评价中的作用。最后，对几种算法的鲁棒性分析表明，所提出的算法具有良好的鲁棒性。

{"title":"Higher-order structure based node importance evaluation in directed networks","authors":"Meng Li , Zhigang Wang , An Zeng , Zengru Di","doi":"10.1016/j.ipm.2024.103948","DOIUrl":"10.1016/j.ipm.2024.103948","url":null,"abstract":"<div><div>Evaluating the significance of objects with possible relevant information is a crucial topic in information science. Due to the fact that objects related to each other can often be described using complex networks, this topic also forms a fundamental theme in network science. Most traditional methods for characterizing the importance of nodes in complex networks only utilize the binary relationships between node pairs, neglecting the influence brought by higher-order structures. Considering the specific interaction modes between local nodes in the network, this paper associates the higher-order structural characteristics of the network with the importance of the nodes. It constructs an evaluation framework for the importance of nodes in directed networks based on higher-order structures. Experimental analysis on both artificial data and scientific citation data from the APS dataset has validated the effectiveness of the proposed algorithms. Compared with PageRank and eigenvector centrality, the proposed algorithms demonstrated higher accuracy, revealing the role of higher-order structures in node importance evaluation. Finally, a robustness analysis of several algorithms indicated that the proposed algorithms exhibited good robustness.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103948"},"PeriodicalIF":7.4,"publicationDate":"2024-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142577974","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multi-view graph contrastive representation learning for bundle recommendation 用于捆绑推荐的多视图对比表示学习

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-11-03 DOI: 10.1016/j.ipm.2024.103956

Peng Zhang , Zhendong Niu , Ru Ma , Fuzhi Zhang

Bundle recommendation can recommend a collection of associated items that can be consumed together to a user rather than recommending these items separately, making it extremely suitable for some scenarios such as product bundle recommendation and game bundle recommendation. Recent bundle recommendation approaches consider auxiliary data to mitigate sparse user-bundle interactions. However, these approaches obtain the node embeddings directly from the established user-bundle graph and do not explicitly exploit the relationships between users (bundles) when constructing recommendation models. Moreover, bundle recommendation approaches based on graph contrastive learning usually construct contrastive views by randomly discarding nodes (edges) in the graph, while discarding some essential nodes or edges will destroy the structure of the original graph, thereby deteriorating the quality of the learned node embeddings. Aiming at these limitations, we propose a bundle recommendation approach based on multi-view graph contrastive representation learning. First, we present a multi-view modeling method to model the relations between entities as several views from different perspectives. These views serve as inputs of graph neural networks for graph representation learning and provide contrastive views for the contrastive learning tasks. Second, we propose a novel framework for bundle recommendation. This framework obtains the user (bundle) embeddings from different views by performing multi-view graph representation learning and enhances the learned user and bundle embeddings through a two-level contrastive learning strategy. On this basis, the enhanced user (bundle) embeddings are fused for prediction. Finally, we design a joint optimization objective to optimize the model parameters, combining the prediction loss that supports multiple negative samples and the contrastive losses. Experiments on the Netease and Youshu datasets reveal that our approach outperforms the state-of-the-art (SOTA) baselines. Furthermore, the average improvements of Recall@K and NDCG@K of our approach over the SOTA baselines are approximately 3.38% and 2.80% on Netease and 3.94% and 4.84% on Youshu.

捆绑推荐可以向用户推荐一系列可以一起消费的关联项目，而不是单独推荐这些项目，因此非常适合一些场景，如产品捆绑推荐和游戏捆绑推荐。最近的捆绑推荐方法考虑了辅助数据，以减轻用户-捆绑交互的稀疏性。然而，这些方法直接从已建立的用户-捆绑图中获取节点嵌入，在构建推荐模型时没有明确利用用户（捆绑）之间的关系。此外，基于图对比学习的捆绑推荐方法通常通过随机丢弃图中的节点（边）来构建对比视图，而丢弃一些重要的节点或边会破坏原始图的结构，从而降低学习到的节点嵌入的质量。针对这些局限性，我们提出了一种基于多视图对比表示学习的捆绑推荐方法。首先，我们提出了一种多视图建模方法，将实体之间的关系建模为来自不同视角的多个视图。这些视图作为图神经网络的输入，用于图表示学习，并为对比学习任务提供对比视图。其次，我们提出了一种新颖的捆绑推荐框架。该框架通过执行多视角图表示学习从不同视角获取用户（捆绑）嵌入，并通过两级对比学习策略增强学习到的用户和捆绑嵌入。在此基础上，融合增强的用户（包）嵌入进行预测。最后，我们设计了一个联合优化目标，结合支持多个负样本的预测损失和对比损失来优化模型参数。在网易和优酷数据集上的实验表明，我们的方法优于最先进的（SOTA）基线。此外，与 SOTA 基线相比，我们的方法在网易数据集上的 Recall@K 和 NDCG@K 平均提高了约 3.38% 和 2.80%，在优树数据集上的 Recall@K 和 NDCG@K 平均提高了约 3.94% 和 4.84%。

{"title":"Multi-view graph contrastive representation learning for bundle recommendation","authors":"Peng Zhang , Zhendong Niu , Ru Ma , Fuzhi Zhang","doi":"10.1016/j.ipm.2024.103956","DOIUrl":"10.1016/j.ipm.2024.103956","url":null,"abstract":"<div><div>Bundle recommendation can recommend a collection of associated items that can be consumed together to a user rather than recommending these items separately, making it extremely suitable for some scenarios such as product bundle recommendation and game bundle recommendation. Recent bundle recommendation approaches consider auxiliary data to mitigate sparse user-bundle interactions. However, these approaches obtain the node embeddings directly from the established user-bundle graph and do not explicitly exploit the relationships between users (bundles) when constructing recommendation models. Moreover, bundle recommendation approaches based on graph contrastive learning usually construct contrastive views by randomly discarding nodes (edges) in the graph, while discarding some essential nodes or edges will destroy the structure of the original graph, thereby deteriorating the quality of the learned node embeddings. Aiming at these limitations, we propose a bundle recommendation approach based on multi-view graph contrastive representation learning. First, we present a multi-view modeling method to model the relations between entities as several views from different perspectives. These views serve as inputs of graph neural networks for graph representation learning and provide contrastive views for the contrastive learning tasks. Second, we propose a novel framework for bundle recommendation. This framework obtains the user (bundle) embeddings from different views by performing multi-view graph representation learning and enhances the learned user and bundle embeddings through a two-level contrastive learning strategy. On this basis, the enhanced user (bundle) embeddings are fused for prediction. Finally, we design a joint optimization objective to optimize the model parameters, combining the prediction loss that supports multiple negative samples and the contrastive losses. Experiments on the Netease and Youshu datasets reveal that our approach outperforms the state-of-the-art (SOTA) baselines. Furthermore, the average improvements of Recall@K and NDCG@K of our approach over the SOTA baselines are approximately 3.38% and 2.80% on Netease and 3.94% and 4.84% on Youshu.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103956"},"PeriodicalIF":7.4,"publicationDate":"2024-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142573158","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Exploring the dynamics of group-based internet rumors propagation: A novel model from the perspective of random hypergraphs 探索基于群体的网络谣言传播动态：随机超图视角下的新型模型

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-11-02 DOI: 10.1016/j.ipm.2024.103941

Yang Xia , Haijun Jiang , Shuzhen Yu

Group interactions have become an important way of online communication today. In this paper, a novel random Hyper-ISDR rumor model is proposed, which uses random hypergraphs to describe the group relationship more accurately. A key innovation of our model is the introduction of hyperpath and path indicators into the group propagation characterization for the first time, explaining the multiple path selectivity present in group propagation. Then, the theoretical conditions for the disappearance and persistence of Internet rumors are obtained by applying stochastic stability theory. This paper finds three interesting results: (1) the propagation threshold on hypergraphs is more sensitive to parameter changes than on traditional graphs; (2) the multiple selectivity of the group propagation path is a critical catalyst for swift rumor diffusion; (3) educating spreaders to become refuters rather than removers is more effective in controlling rumors. Moreover, compared with the graph-based ISDR model and the Hyper-SIR model, it shows that the hyperdegree and path indicators have a greater impact on rumor volatility. Finally, the reliability and applicability of the results are verified by numerical simulation and a real-life case study. This work not only opens up a new perspective of group rumor dynamics analysis, but also provides a superior framework for understanding and managing online information diffusion.

群体互动已成为当今网络交流的一种重要方式。本文提出了一种新颖的随机超ISDR谣言模型，利用随机超图来更准确地描述群组关系。我们模型的一个重要创新是首次将超路径和路径指标引入群传播表征，解释了群传播中存在的多路径选择性。然后，运用随机稳定性理论得到了网络谣言消失和持续存在的理论条件。本文发现了三个有趣的结果：(1) 与传统图相比，超图上的传播阈值对参数变化更加敏感；(2) 群传播路径的多重选择性是谣言迅速扩散的关键催化剂；(3) 教育传播者成为反驳者而非清除者更能有效控制谣言。此外，与基于图的 ISDR 模型和 Hyper-SIR 模型相比，研究表明超度指标和路径指标对谣言波动性的影响更大。最后，通过数值模拟和实际案例研究验证了结果的可靠性和适用性。这项工作不仅开辟了群体谣言动态分析的新视角，而且为理解和管理网络信息扩散提供了一个卓越的框架。

{"title":"Exploring the dynamics of group-based internet rumors propagation: A novel model from the perspective of random hypergraphs","authors":"Yang Xia , Haijun Jiang , Shuzhen Yu","doi":"10.1016/j.ipm.2024.103941","DOIUrl":"10.1016/j.ipm.2024.103941","url":null,"abstract":"<div><div>Group interactions have become an important way of online communication today. In this paper, a novel random Hyper-ISDR rumor model is proposed, which uses random hypergraphs to describe the group relationship more accurately. A key innovation of our model is the introduction of hyperpath and path indicators into the group propagation characterization for the first time, explaining the multiple path selectivity present in group propagation. Then, the theoretical conditions for the disappearance and persistence of Internet rumors are obtained by applying stochastic stability theory. This paper finds three interesting results: (1) the propagation threshold on hypergraphs is more sensitive to parameter changes than on traditional graphs; (2) the multiple selectivity of the group propagation path is a critical catalyst for swift rumor diffusion; (3) educating spreaders to become refuters rather than removers is more effective in controlling rumors. Moreover, compared with the graph-based ISDR model and the Hyper-SIR model, it shows that the hyperdegree and path indicators have a greater impact on rumor volatility. Finally, the reliability and applicability of the results are verified by numerical simulation and a real-life case study. This work not only opens up a new perspective of group rumor dynamics analysis, but also provides a superior framework for understanding and managing online information diffusion.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103941"},"PeriodicalIF":7.4,"publicationDate":"2024-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142573156","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An optimal multi-scale and multi-factor two-stage integration paradigm coupled with investor sentiment for carbon price prediction 结合投资者情绪的多尺度、多因素两阶段优化整合范式，用于碳价格预测

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-11-02 DOI: 10.1016/j.ipm.2024.103953

Jujie Wang, Xuecheng He

The accurate prediction of carbon emission trading prices is of great significance for the effective allocation of carbon resources, achieving energy conservation, emission reduction, and green development. However, it is difficult to fully extract the fluctuation information of carbon price, and external factors also have complex impacts on it, so it is a challenge to accurately predict carbon price. Therefore, this study proposes an optimal multi-scale and multi-factor two-stage integration paradigm coupled with investor sentiment for carbon price prediction. Firstly, an adaptive periodic variational mode decomposition (APVMD) method is proposed to capture feature subsequences with different fluctuation information from a periodic perspective in carbon prices. Then a comprehensive impact factor library is constructed to assist in prediction, including unstructured data on investor sentiment and structured data. Through the enhanced light gradient boosting machine (ELightGBM) algorithm, the optimal driving factors for each feature subsequence are fully screened, and the dimensionality of the data is reduced based on their nonlinear relationship. Considering the selection of hyperparameters and the contribution of different feature subsequences, an optimized two-stage integrated prediction is designed to achieve high-precision point prediction. On this basis, uncertainty analysis is used to obtain reasonable interval prediction results. Through comparative analysis, this model is better than other comparative models in terms of predictive ability and stability.

准确预测碳排放权交易价格对有效配置碳资源、实现节能减排和绿色发展具有重要意义。然而，碳价格的波动信息难以完全提取，外部因素对其影响也较为复杂，因此准确预测碳价格是一项挑战。因此，本研究提出了一种与投资者情绪相结合的多尺度、多因素两阶段最优集成范式，用于碳价格预测。首先，提出一种自适应周期变异模式分解（APVMD）方法，从周期性角度捕捉碳价格中具有不同波动信息的特征子序列。然后，构建了一个全面的影响因子库来辅助预测，包括投资者情绪的非结构化数据和结构化数据。通过增强型光梯度提升机（ELightGBM）算法，充分筛选出各特征子序列的最优驱动因子，并根据其非线性关系降低数据维度。考虑到超参数的选择和不同特征子序列的贡献，设计了优化的两阶段综合预测，以实现高精度的点预测。在此基础上，利用不确定性分析获得合理的区间预测结果。通过比较分析，该模型在预测能力和稳定性方面优于其他比较模型。

{"title":"An optimal multi-scale and multi-factor two-stage integration paradigm coupled with investor sentiment for carbon price prediction","authors":"Jujie Wang, Xuecheng He","doi":"10.1016/j.ipm.2024.103953","DOIUrl":"10.1016/j.ipm.2024.103953","url":null,"abstract":"<div><div>The accurate prediction of carbon emission trading prices is of great significance for the effective allocation of carbon resources, achieving energy conservation, emission reduction, and green development. However, it is difficult to fully extract the fluctuation information of carbon price, and external factors also have complex impacts on it, so it is a challenge to accurately predict carbon price. Therefore, this study proposes an optimal multi-scale and multi-factor two-stage integration paradigm coupled with investor sentiment for carbon price prediction. Firstly, an adaptive periodic variational mode decomposition (APVMD) method is proposed to capture feature subsequences with different fluctuation information from a periodic perspective in carbon prices. Then a comprehensive impact factor library is constructed to assist in prediction, including unstructured data on investor sentiment and structured data. Through the enhanced light gradient boosting machine (ELightGBM) algorithm, the optimal driving factors for each feature subsequence are fully screened, and the dimensionality of the data is reduced based on their nonlinear relationship. Considering the selection of hyperparameters and the contribution of different feature subsequences, an optimized two-stage integrated prediction is designed to achieve high-precision point prediction. On this basis, uncertainty analysis is used to obtain reasonable interval prediction results. Through comparative analysis, this model is better than other comparative models in terms of predictive ability and stability.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103953"},"PeriodicalIF":7.4,"publicationDate":"2024-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142573157","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Sequential recommendation by reprogramming pretrained transformer 通过重新编程预训练变换器进行顺序推荐

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-11-01 DOI: 10.1016/j.ipm.2024.103938

Min Tang , Shujie Cui , Zhe Jin , Shiuan-ni Liang , Chenliang Li , Lixin Zou

Inspired by the success of Pre-trained language models (PLMs), numerous sequential recommenders attempted to replicate its achievements by employing PLMs’ efficient architectures for building large models and using self-supervised learning for broadening training data. Despite their success, there is curiosity about developing a large-scale sequential recommender system since existing methods either build models within a single dataset or utilize text as an intermediary for alignment across different datasets. However, due to the sparsity of user–item interactions, unalignment between different datasets, and lack of global information in the sequential recommendation, directly pre-training a large foundation model may not be feasible.

Towards this end, we propose the RecPPT that firstly employs the GPT-2 to model historical sequence by training the input item embedding and the output layer from scratch, which avoids training a large model on the sparse user–item interactions. Additionally, to alleviate the burden of unalignment, the RecPPT is equipped with a reprogramming module to reprogram the target embedding to existing well-trained proto-embeddings. Furthermore, RecPPT integrates global information into sequences by initializing the item embedding using an SVD-based initializer. Extensive experiments over four datasets demonstrated the RecPPT achieved an average improvement of 6.5% on NDCG@5, 6.2% on NDCG@10, 6.1% on Recall@5, and 5.4% on Recall@10 compared to the baselines. Particularly in few-shot scenarios, the significant improvements in NDCG@10 confirm the superiority of the proposed method.

受到预训练语言模型（PLMs）成功的启发，许多顺序推荐系统试图复制其成就，方法是采用预训练语言模型的高效架构来构建大型模型，并利用自监督学习来扩大训练数据。尽管取得了成功，但人们对开发大规模顺序推荐系统仍充满好奇，因为现有的方法要么是在单个数据集内建立模型，要么是利用文本作为中介在不同数据集之间进行排列。为此，我们提出了 RecPPT，首先利用 GPT-2 建立历史序列模型，从头开始训练输入项嵌入和输出层，从而避免在稀疏的用户-项交互上训练大型模型。此外，为了减轻不对齐的负担，RecPPT 还配备了一个重新编程模块，可根据现有训练有素的原嵌入对目标嵌入进行重新编程。此外，RecPPT 还使用基于 SVD 的初始化器初始化项目嵌入，从而将全局信息整合到序列中。在四个数据集上进行的广泛实验表明，与基线相比，RecPPT 在 NDCG@5 上平均提高了 6.5%，在 NDCG@10 上平均提高了 6.2%，在 Recall@5 上平均提高了 6.1%，在 Recall@10 上平均提高了 5.4%。特别是在拍摄次数较少的情况下，NDCG@10 的显著提高证实了所提方法的优越性。

{"title":"Sequential recommendation by reprogramming pretrained transformer","authors":"Min Tang , Shujie Cui , Zhe Jin , Shiuan-ni Liang , Chenliang Li , Lixin Zou","doi":"10.1016/j.ipm.2024.103938","DOIUrl":"10.1016/j.ipm.2024.103938","url":null,"abstract":"<div><div>Inspired by the success of Pre-trained language models (PLMs), numerous sequential recommenders attempted to replicate its achievements by employing PLMs’ efficient architectures for building large models and using self-supervised learning for broadening training data. Despite their success, there is curiosity about developing a large-scale sequential recommender system since existing methods either build models within a single dataset or utilize text as an intermediary for alignment across different datasets. However, due to the sparsity of user–item interactions, unalignment between different datasets, and lack of global information in the sequential recommendation, directly pre-training a large foundation model may not be feasible.</div><div>Towards this end, we propose the <span>RecPPT</span> that firstly employs the GPT-2 to model historical sequence by training the input item embedding and the output layer from scratch, which avoids training a large model on the sparse user–item interactions. Additionally, to alleviate the burden of unalignment, the <span>RecPPT</span> is equipped with a reprogramming module to reprogram the target embedding to existing well-trained proto-embeddings. Furthermore, <span>RecPPT</span> integrates global information into sequences by initializing the item embedding using an SVD-based initializer. Extensive experiments over four datasets demonstrated the <span>RecPPT</span> achieved an average improvement of 6.5% on NDCG@5, 6.2% on NDCG@10, 6.1% on Recall@5, and 5.4% on Recall@10 compared to the baselines. Particularly in few-shot scenarios, the significant improvements in NDCG@10 confirm the superiority of the proposed method.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103938"},"PeriodicalIF":7.4,"publicationDate":"2024-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142573227","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Moment matching of joint distributions for unsupervised domain adaptation 用于无监督领域适应的联合分布矩匹配

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-10-31 DOI: 10.1016/j.ipm.2024.103944

Bo Zhang , Xiaoming Zhang , Zhibo Zhou , Yun Liu , Yancong Li , Feiran Huang

Unsupervised Domain Adaptation (UDA) is designed to transfer acquired knowledge from the source domain to an unlabeled target domain. In this paper, we present a comprehensive approach that seamlessly addresses both source-available and source-free UDA by matching the joint distributions across domains, independent of the availability of source data. Our methodology introduces three innovative criteria to quantitatively assess the divergences between the source and target data, as well as between the source model hypothesis and target data. The criteria decide whether the predicted labels of the target hypothesis are affected by the other knowledge of both domains in the form of a precise formula, thereby enabling targeted supervision in UDA. We evaluate the effectiveness through 37 image and text classification tasks across four different datasets, comparing their performance against the state-of-the-art models. Experiments demonstrate that the proposed approaches obtain superior accuracies for most of the tasks, especially for the source-free setting, which still exceeds HOMDA 0.6% on Office and DRDA 1.5% on Office-Home, even without direct access to source data.

无监督领域适应（UDA）旨在将源领域中获得的知识转移到未标记的目标领域中。在本文中，我们提出了一种综合方法，通过匹配跨域的联合分布，无缝地解决了有源和无源 UDA 的问题，而与源数据的可用性无关。我们的方法引入了三个创新标准，用于定量评估源数据和目标数据之间的差异，以及源模型假设和目标数据之间的差异。这些标准以精确公式的形式决定目标假设的预测标签是否受到两个领域其他知识的影响，从而在 UDA 中实现有针对性的监督。我们通过四个不同数据集的 37 项图像和文本分类任务来评估其有效性，并将其性能与最先进的模型进行比较。实验表明，所提出的方法在大多数任务中都获得了优异的准确度，尤其是在无源设置中，即使不直接访问源数据，在 Office 和 DRDA 中的准确度仍分别超过 HOMDA 的 0.6% 和 DRDA 的 1.5%。

{"title":"Moment matching of joint distributions for unsupervised domain adaptation","authors":"Bo Zhang , Xiaoming Zhang , Zhibo Zhou , Yun Liu , Yancong Li , Feiran Huang","doi":"10.1016/j.ipm.2024.103944","DOIUrl":"10.1016/j.ipm.2024.103944","url":null,"abstract":"<div><div>Unsupervised Domain Adaptation (UDA) is designed to transfer acquired knowledge from the source domain to an unlabeled target domain. In this paper, we present a comprehensive approach that seamlessly addresses both source-available and source-free UDA by matching the joint distributions across domains, independent of the availability of source data. Our methodology introduces three innovative criteria to quantitatively assess the divergences between the source and target data, as well as between the source model hypothesis and target data. The criteria decide whether the predicted labels of the target hypothesis are affected by the other knowledge of both domains in the form of a precise formula, thereby enabling targeted supervision in UDA. We evaluate the effectiveness through 37 image and text classification tasks across four different datasets, comparing their performance against the state-of-the-art models. Experiments demonstrate that the proposed approaches obtain superior accuracies for most of the tasks, especially for the source-free setting, which still exceeds HOMDA 0.6% on Office and DRDA 1.5% on Office-Home, even without direct access to source data.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103944"},"PeriodicalIF":7.4,"publicationDate":"2024-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142553583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Quantity forecast of mobile subscribers with Time-Dilated Attention 时间稀释注意力的移动用户数量预测

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-10-30 DOI: 10.1016/j.ipm.2024.103940

Binhong Yao

The quantity forecast of mobile subscribers requires accurate and reliable results for obtaining insights into user trends and facilitating effective business management. Due to the complexity inherent in mobile subscriber data, influenced by subscriber tendencies and device popularity, capturing its underlying regularities poses a challenge. In this research, a novel Time-Dilated Attention (TDA) model is proposed, complemented by a feature extraction method characterized by high interpretability and distinguishability. Its efficacy and implications are explored on a real-world mobile subscriber dataset. TDA facilitates the acquisition of more informative representations, while our feature extraction method enhances the ability to discern dissimilar samples, thereby improving the stability of mobile subscriber trend analysis. The approach is validated on three additional datasets to assess its robustness. Experimental findings on the target mobile subscriber dataset demonstrate that the proposed approach achieves reductions in MAE, RMSE, and Theil’s U by 1.45%, 5.28%, and 5.12%, respectively, compared to the strongest baseline methods. Additionally, it attains the second-best performance in terms of MedAE. Furthermore, this model consistently ranks within the top two positions for nine out of twelve metrics on the additional datasets, underscoring its generalizability.

移动用户数量预测需要准确可靠的结果，以便深入了解用户趋势，促进有效的业务管理。由于移动用户数据固有的复杂性，受用户趋势和设备流行度的影响，捕捉其潜在的规律性是一项挑战。在这项研究中，提出了一种新颖的时间稀释注意力（TDA）模型，并辅以一种具有高度可解释性和可区分性的特征提取方法。研究人员在真实世界的移动用户数据集上探索了该模型的功效和意义。TDA 有助于获取更多的信息表征，而我们的特征提取方法则增强了对不同样本的辨别能力，从而提高了移动用户趋势分析的稳定性。该方法在另外三个数据集上进行了验证，以评估其鲁棒性。在目标移动用户数据集上的实验结果表明，与最强的基线方法相比，所提出的方法在 MAE、RMSE 和 Theil's U 方面分别降低了 1.45%、5.28% 和 5.12%。此外，它在 MedAE 方面的表现也是第二好的。此外，在其他数据集的 12 项指标中，该模型有 9 项指标始终保持在前两名的位置，这突出表明了它的通用性。

{"title":"Quantity forecast of mobile subscribers with Time-Dilated Attention","authors":"Binhong Yao","doi":"10.1016/j.ipm.2024.103940","DOIUrl":"10.1016/j.ipm.2024.103940","url":null,"abstract":"<div><div>The quantity forecast of mobile subscribers requires accurate and reliable results for obtaining insights into user trends and facilitating effective business management. Due to the complexity inherent in mobile subscriber data, influenced by subscriber tendencies and device popularity, capturing its underlying regularities poses a challenge. In this research, a novel Time-Dilated Attention (TDA) model is proposed, complemented by a feature extraction method characterized by high interpretability and distinguishability. Its efficacy and implications are explored on a real-world mobile subscriber dataset. TDA facilitates the acquisition of more informative representations, while our feature extraction method enhances the ability to discern dissimilar samples, thereby improving the stability of mobile subscriber trend analysis. The approach is validated on three additional datasets to assess its robustness. Experimental findings on the target mobile subscriber dataset demonstrate that the proposed approach achieves reductions in MAE, RMSE, and Theil’s U by 1.45%, 5.28%, and 5.12%, respectively, compared to the strongest baseline methods. Additionally, it attains the second-best performance in terms of MedAE. Furthermore, this model consistently ranks within the top two positions for nine out of twelve metrics on the additional datasets, underscoring its generalizability.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103940"},"PeriodicalIF":7.4,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142553582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Spatial and temporal twin-guided pattern recurrent graph network for implementing reasoning of spatiotemporal knowledge graph 用于实现时空知识图谱推理的时空孪生引导模式循环图网络

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-10-30 DOI: 10.1016/j.ipm.2024.103942

Xiaobei Xu , Ruizhe Ma , Beijing Zhou , Li Yan , Zongmin Ma

The extrapolation of knowledge graphs (KGs) has been the subject of numerous studies. However, real world data often has complex spatial attributes, which makes reasoning on spatiotemporal knowledge graphs (STKGs) challenging. In response, we propose a model that captures both temporal and spatial patterns to address the challenge of predicting future facts in STKGs. The proposed spatial and temporal twin-guided pattern recurrent graph network (STTP-RGN) utilizes temporal and spatial sequences to identify cyclic and repetitive patterns in data. It performs spatiotemporal-twin encoding and temporal and spatial sequence encoding respectively, and inputs the encoded three results into three corresponding decoders to determine the evolution of entity and predicate representations in time and space. We used the YAGO10K, Wikidata40K, Opensky18K and DY-NB21K for tests on entity and predicate prediction. On YAGO10K, the model's entity prediction performance outperforms the best temporal extrapolation model RETIA by 20 %. The predicate and entity predictions on Wikidata40K have improved by 3 % and 20 %, respectively. Results for entity prediction on Opensky18K have increased by 30 %, while results for predicate prediction have improved by 1 %. The experimental results demonstrate that the model fills the gap in knowledge extrapolation on STKG.

知识图谱（KG）的外推一直是众多研究的主题。然而，现实世界的数据往往具有复杂的空间属性，这使得时空知识图谱（STKGs）推理具有挑战性。为此，我们提出了一种同时捕捉时间和空间模式的模型，以应对在 STKGs 中预测未来事实的挑战。我们提出的时空孪生引导模式循环图网络（STTP-RGN）利用时间和空间序列来识别数据中的循环和重复模式。它分别执行时空孪生编码和时空序列编码，并将编码后的三个结果输入三个相应的解码器，以确定实体和谓词表示在时间和空间上的演变。我们使用 YAGO10K、Wikidata40K、Opensky18K 和 DY-NB21K 对实体和谓词预测进行了测试。在 YAGO10K 上，该模型的实体预测性能比最佳时间外推模型 RETIA 高出 20%。在 Wikidata40K 上，谓词和实体预测分别提高了 3% 和 20%。Opensky18K 上的实体预测结果提高了 30%，而谓词预测结果提高了 1%。实验结果表明，该模型填补了 STKG 知识外推方面的空白。

{"title":"Spatial and temporal twin-guided pattern recurrent graph network for implementing reasoning of spatiotemporal knowledge graph","authors":"Xiaobei Xu , Ruizhe Ma , Beijing Zhou , Li Yan , Zongmin Ma","doi":"10.1016/j.ipm.2024.103942","DOIUrl":"10.1016/j.ipm.2024.103942","url":null,"abstract":"<div><div>The extrapolation of knowledge graphs (KGs) has been the subject of numerous studies. However, real world data often has complex spatial attributes, which makes reasoning on spatiotemporal knowledge graphs (STKGs) challenging. In response, we propose a model that captures both temporal and spatial patterns to address the challenge of predicting future facts in STKGs. The proposed spatial and temporal twin-guided pattern recurrent graph network (STTP-RGN) utilizes temporal and spatial sequences to identify cyclic and repetitive patterns in data. It performs spatiotemporal-twin encoding and temporal and spatial sequence encoding respectively, and inputs the encoded three results into three corresponding decoders to determine the evolution of entity and predicate representations in time and space. We used the YAGO10K, Wikidata40K, Opensky18K and DY-NB21K for tests on entity and predicate prediction. On YAGO10K, the model's entity prediction performance outperforms the best temporal extrapolation model RETIA by 20 %. The predicate and entity predictions on Wikidata40K have improved by 3 % and 20 %, respectively. Results for entity prediction on Opensky18K have increased by 30 %, while results for predicate prediction have improved by 1 %. The experimental results demonstrate that the model fills the gap in knowledge extrapolation on STKG.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103942"},"PeriodicalIF":7.4,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142553495","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A comprehensive survey on social engineering attacks, countermeasures, case study, and research challenges 关于社会工程学攻击、对策、案例研究和研究挑战的全面调查

IF 7.4 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2024-10-30 DOI: 10.1016/j.ipm.2024.103928

Tejal Rathod , Nilesh Kumar Jadav , Sudeep Tanwar , Abdulatif Alabdulatif , Deepak Garg , Anupam Singh

Social engineering attacks are inevitable and imperil the integrity, security, and confidentiality of the information used on social media platforms. Prominent technologies, such as blockchain, artificial intelligence (AI), and proactive access controls, were adopted in the literature to confront the social engineering attacks on social media. Nevertheless, a comprehensive survey on this topic is notably absent from the current body of research. Inspired by that, we propose an exhaustive survey comprising an in-depth analysis of 10 distinct social engineering attacks with their real-time scenarios. Furthermore, a detailed solution taxonomy is presented, offering valuable insights (e.g., objective, methodology, and results) to tackle social engineering attacks effectively. Based on the solution taxonomy, we propose an AI and blockchain-based malicious uniform resource locator (URL) detection framework (as a case study) to confront social engineering attacks on the Meta platform. For that, a standard dataset is utilized, which comprises 12 different datasets containing 3980870 malicious and non-malicious URLs. To classify URLs, a binary classification problem is formulated and solved by using different AI classifiers, such as Naive Bayes (NB), decision tree (DT), support vector machine (SVM), and boosted tree (BT). The non-malicious URLs are forwarded to the blockchain network to ensure secure storage, strengthening the effectiveness of the malicious URL detection framework. The proposed framework is evaluated with baseline approaches, wherein the NB achieves noteworthy training accuracy, i.e., 76.87% and training time of (8.23 (s)). Additionally, interplanetary file system (IPFS)-based blockchain achieves a remarkable response time, i.e., (0.245 (ms)) compared to the conventional blockchain technology. We also used execution cost and smart contract vulnerability assessment using Slither to showcase the outperformance of blockchain technology. Lastly, we shed light on the open issues and research challenges of social engineering attacks where research gaps still exist and require further investigation.

社交工程攻击是不可避免的，它危及社交媒体平台上所使用信息的完整性、安全性和保密性。文献中采用了区块链、人工智能（AI）和主动访问控制等著名技术来应对社交媒体上的社交工程攻击。然而，目前的研究成果中明显缺乏对这一主题的全面调查。受此启发，我们提出了一份详尽的调查报告，其中包括对 10 种不同社交工程攻击及其实时场景的深入分析。此外，我们还提出了详细的解决方案分类法，为有效解决社会工程学攻击提供了有价值的见解（如目标、方法和结果）。基于解决方案分类法，我们提出了一个基于人工智能和区块链的恶意统一资源定位器（URL）检测框架（作为案例研究），以应对 Meta 平台上的社交工程攻击。为此，我们使用了一个标准数据集，其中包括 12 个不同的数据集，包含 3980870 个恶意和非恶意 URL。为了对 URL 进行分类，制定了一个二元分类问题，并使用不同的人工智能分类器（如 Naive Bayes (NB)、决策树 (DT)、支持向量机 (SVM) 和助推树 (BT)）加以解决。非恶意 URL 被转发到区块链网络以确保安全存储，从而加强了恶意 URL 检测框架的有效性。所提出的框架与基线方法进行了评估，其中 NB 的训练准确率达到了值得注意的水平，即 76.87%，训练时间为（8.23 (s)）。此外，与传统的区块链技术相比，基于星际文件系统（IPFS）的区块链实现了显著的响应时间，即（0.245 (ms)）。我们还利用 Slither 进行了执行成本和智能合约漏洞评估，以展示区块链技术的优越性能。最后，我们揭示了社会工程学攻击的公开问题和研究挑战，这些问题和挑战仍存在研究空白，需要进一步研究。

{"title":"A comprehensive survey on social engineering attacks, countermeasures, case study, and research challenges","authors":"Tejal Rathod , Nilesh Kumar Jadav , Sudeep Tanwar , Abdulatif Alabdulatif , Deepak Garg , Anupam Singh","doi":"10.1016/j.ipm.2024.103928","DOIUrl":"10.1016/j.ipm.2024.103928","url":null,"abstract":"<div><div>Social engineering attacks are inevitable and imperil the integrity, security, and confidentiality of the information used on social media platforms. Prominent technologies, such as blockchain, artificial intelligence (AI), and proactive access controls, were adopted in the literature to confront the social engineering attacks on social media. Nevertheless, a comprehensive survey on this topic is notably absent from the current body of research. Inspired by that, we propose an exhaustive survey comprising an in-depth analysis of 10 distinct social engineering attacks with their real-time scenarios. Furthermore, a detailed solution taxonomy is presented, offering valuable insights (e.g., objective, methodology, and results) to tackle social engineering attacks effectively. Based on the solution taxonomy, we propose an AI and blockchain-based malicious uniform resource locator (URL) detection framework (as a case study) to confront social engineering attacks on the Meta platform. For that, a standard dataset is utilized, which comprises 12 different datasets containing 3980870 malicious and non-malicious URLs. To classify URLs, a binary classification problem is formulated and solved by using different AI classifiers, such as Naive Bayes (NB), decision tree (DT), support vector machine (SVM), and boosted tree (BT). The non-malicious URLs are forwarded to the blockchain network to ensure secure storage, strengthening the effectiveness of the malicious URL detection framework. The proposed framework is evaluated with baseline approaches, wherein the NB achieves noteworthy training accuracy, i.e., 76.87% and training time of (8.23 (s)). Additionally, interplanetary file system (IPFS)-based blockchain achieves a remarkable response time, i.e., (0.245 (ms)) compared to the conventional blockchain technology. We also used execution cost and smart contract vulnerability assessment using Slither to showcase the outperformance of blockchain technology. Lastly, we shed light on the open issues and research challenges of social engineering attacks where research gaps still exist and require further investigation.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 1","pages":"Article 103928"},"PeriodicalIF":7.4,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142553496","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0