Information Processing & Management最新文献_第7页

Enhancing large language models for knowledge graph question answering via multi-granularity knowledge injection and structured reasoning path-augmented prompting 通过多粒度知识注入和结构化推理路径增强提示增强知识图问答的大型语言模型

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-13 DOI: 10.1016/j.ipm.2026.104614

Chuanyang Gong , Zhihua Wei , Wenhao Tao , Duoqian Miao

Large language models (LLMs) often exhibit factual errors when handling complex knowledge reasoning. To address this issue, we propose MGPrompt, a novel knowledge graph question answering (KGQA) framework that enhances LLM performance by integrating multi-granularity knowledge with structured reasoning path-augmented prompting. MGPrompt consists of three core modules-knowledge refinement, semantic association, and information fusion-to dynamically filter and integrate entity-level, relation-level, and subgraph-level knowledge retrieved from the knowledge graph. Subsequently, we inject these refined semantic representations as prefix vectors into the LLM and fine-tune the model using Low-Rank Adaptation (LoRA) to guide it in generating accurate reasoning paths. We conducted extensive experiments on two benchmark datasets, WebQSP and CWQ. The results show that MGPrompt achieves highly competitive performance compared to 30 baseline methods. Experimental results show that MGPrompt achieves highly competitive performance on both WebQSP and CWQ; in particular, it improves the Hits@1 score on WebQSP by 1.1% over the strongest baseline (85.7%), thereby clearly demonstrating the effectiveness of the proposed framework for complex KGQA tasks.

大型语言模型（llm）在处理复杂的知识推理时经常出现事实错误。为了解决这个问题，我们提出了MGPrompt，这是一个新的知识图问答（KGQA）框架，通过将多粒度知识与结构化推理路径增强提示集成在一起来提高LLM的性能。MGPrompt由知识精化、语义关联和信息融合三个核心模块组成，用于动态过滤和集成从知识图中检索到的实体级、关系级和子图级知识。随后，我们将这些精炼的语义表示作为前缀向量注入到LLM中，并使用低秩自适应（Low-Rank Adaptation, LoRA）对模型进行微调，引导其生成准确的推理路径。我们在WebQSP和CWQ两个基准数据集上进行了大量的实验。结果表明，与30种基准方法相比，MGPrompt实现了极具竞争力的性能。实验结果表明，MGPrompt在WebQSP和CWQ上都具有很强的竞争力；特别是，它在WebQSP上的Hits@1得分比最强基线（85.7%）提高了1.1%，从而清楚地证明了所提出的框架对于复杂的KGQA任务的有效性。

{"title":"Enhancing large language models for knowledge graph question answering via multi-granularity knowledge injection and structured reasoning path-augmented prompting","authors":"Chuanyang Gong , Zhihua Wei , Wenhao Tao , Duoqian Miao","doi":"10.1016/j.ipm.2026.104614","DOIUrl":"10.1016/j.ipm.2026.104614","url":null,"abstract":"<div><div>Large language models (LLMs) often exhibit factual errors when handling complex knowledge reasoning. To address this issue, we propose MGPrompt, a novel knowledge graph question answering (KGQA) framework that enhances LLM performance by integrating multi-granularity knowledge with structured reasoning path-augmented prompting. MGPrompt consists of three core modules-knowledge refinement, semantic association, and information fusion-to dynamically filter and integrate entity-level, relation-level, and subgraph-level knowledge retrieved from the knowledge graph. Subsequently, we inject these refined semantic representations as prefix vectors into the LLM and fine-tune the model using Low-Rank Adaptation (LoRA) to guide it in generating accurate reasoning paths. We conducted extensive experiments on two benchmark datasets, WebQSP and CWQ. The results show that MGPrompt achieves highly competitive performance compared to 30 baseline methods. Experimental results show that MGPrompt achieves highly competitive performance on both WebQSP and CWQ; in particular, it improves the Hits@1 score on WebQSP by 1.1% over the strongest baseline (85.7%), thereby clearly demonstrating the effectiveness of the proposed framework for complex KGQA tasks.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104614"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978423","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

NeuPath: A hybrid learning-based optimization approach for emergency search path planning NeuPath：一种基于混合学习的紧急搜索路径规划优化方法

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-14 DOI: 10.1016/j.ipm.2026.104615

Yingying Gao , Tianle Pu , Qingqing Yang , Zhiwei Yang , Kewei Yang , Changjun Fan

Search operations play a vital role in humanitarian emergencies, where maximizing the probability of finding survivors requires highly efficient solutions. A key challenge lies in the limitations of existing methods, which often struggle with high-dimensional constraints and sparse decision spaces, particularly in large-scale scenarios. To address this, we propose NeuPath, a hybrid learning-based optimization framework that accelerates the discovery of high-quality solutions. NeuPath first formulates the optimal search problem (OSP) as a bipartite graph representation to enhance feature extraction and scalability. It then predicts an initial solution using a graph neural network (GNN) augmented with a two-stage aggregation mechanism, followed by refinement via a block-wise trust-region scheme. Extensive experiments on OSP static scenarios (500 instances) demonstrate that NeuPath achieves significant speedups over exact solvers, with performance gains of 2.48 ×  (Gurobi) and 2.74 ×  (SCIP) across varying problem sizes. For large-scale random scenarios (500 instances), the solution quality of this method also significantly exceeds that of the exact solver in a finite time (3600s). Moreover, the framework exhibits strong generalization capabilities by learning meaningful problem structure features. Ablation studies further validate the effectiveness of each module.

搜索行动在人道主义紧急情况中发挥着至关重要的作用，在这种情况下，最大限度地找到幸存者的可能性需要高效的解决方案。一个关键的挑战在于现有方法的局限性，这些方法经常与高维约束和稀疏决策空间作斗争，特别是在大规模场景中。为了解决这个问题，我们提出了NeuPath，这是一个基于学习的混合优化框架，可以加速发现高质量的解决方案。NeuPath首先将最优搜索问题（OSP）表述为二部图表示，以增强特征提取和可扩展性。然后，它使用带有两阶段聚合机制的图神经网络（GNN）预测初始解决方案，然后通过块方向的信任区域方案进行细化。在OSP静态场景（500个实例）上进行的大量实验表明，NeuPath比精确求解器实现了显著的加速，在不同的问题规模上，性能提高了2.48 × （Gurobi）和2.74 × （SCIP）。对于大规模随机场景（500个实例），该方法在有限时间（3600s）内的求解质量也明显优于精确求解器。此外，该框架通过学习有意义的问题结构特征，表现出较强的泛化能力。烧蚀研究进一步验证了各模块的有效性。

{"title":"NeuPath: A hybrid learning-based optimization approach for emergency search path planning","authors":"Yingying Gao , Tianle Pu , Qingqing Yang , Zhiwei Yang , Kewei Yang , Changjun Fan","doi":"10.1016/j.ipm.2026.104615","DOIUrl":"10.1016/j.ipm.2026.104615","url":null,"abstract":"<div><div>Search operations play a vital role in humanitarian emergencies, where maximizing the probability of finding survivors requires highly efficient solutions. A key challenge lies in the limitations of existing methods, which often struggle with high-dimensional constraints and sparse decision spaces, particularly in large-scale scenarios. To address this, we propose NeuPath, a hybrid learning-based optimization framework that accelerates the discovery of high-quality solutions. NeuPath first formulates the optimal search problem (OSP) as a bipartite graph representation to enhance feature extraction and scalability. It then predicts an initial solution using a graph neural network (GNN) augmented with a two-stage aggregation mechanism, followed by refinement via a block-wise trust-region scheme. Extensive experiments on OSP static scenarios (500 instances) demonstrate that NeuPath achieves significant speedups over exact solvers, with performance gains of 2.48 ×  (Gurobi) and 2.74 ×  (SCIP) across varying problem sizes. For large-scale random scenarios (500 instances), the solution quality of this method also significantly exceeds that of the exact solver in a finite time (3600s). Moreover, the framework exhibits strong generalization capabilities by learning meaningful problem structure features. Ablation studies further validate the effectiveness of each module.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104615"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978434","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

Cross-modal information propagation for contrastive multi-modal clustering 对比多模态聚类的跨模态信息传播

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-06 DOI: 10.1016/j.ipm.2025.104595

Tongji Chen , Guoliang Zou , Shizhe Hu, Yangdong Ye

Multi-modal clustering aims to exploit relationships between different modalities to enhance clustering performance. However, existing methods face two main challenges. First, feature extraction fails to fully utilize the relationships between samples from different modalities, which is crucial for capturing global multi-modal information. Second, most clustering methods struggle to correct significantly erroneous assignments. To address these challenges, we propose Cross-modal Information Propagation for Contrastive Multi-modal Clustering (CIPCMC), a novel method driven by cross-modal information (CMI) and contrastive learning. We progressively obtain private CMI and integrate it into a unified CMI, which is then propagated to optimize the entire model. First, a cross-attention mechanism introduces CMI for each modality, enabling the model to focus on relationships between different modalities. This allows the model to uncover semantic associations and effectively exploit the complementary nature of multi-modal data. Next, we fuse modality-specific representations to derive a unified CMI representation, which helps each modality correct erroneous assignments, leading to high-confidence clustering. The end-to-end training of CIPCMC ensures module synergy, improving performance and generalization. Experiments on challenging datasets show that CIPCMC outperforms existing methods, achieving accuracy improvements of 10.0% on the Caltech-3M dataset and 16.6% on the PBMC dataset.

多模态聚类旨在利用不同模态之间的关系来提高聚类性能。然而，现有的方法面临两个主要挑战。首先，特征提取未能充分利用不同模态样本之间的关系，而这对于捕获全局多模态信息至关重要。其次，大多数聚类方法都难以纠正明显错误的分配。为了解决这些挑战，我们提出了一种由跨模态信息（CMI）和对比学习驱动的跨模态信息传播的对比多模态聚类方法（CIPCMC）。我们逐步得到私有CMI，并将其整合成一个统一的CMI，然后将其传播以优化整个模型。首先，交叉注意机制为每个模态引入CMI，使模型能够关注不同模态之间的关系。这使得模型能够发现语义关联，并有效地利用多模态数据的互补性。接下来，我们融合特定于模态的表示来获得统一的CMI表示，这有助于每个模态纠正错误分配，从而实现高置信度聚类。CIPCMC的端到端培训确保了模块的协同，提高了性能和通用性。在具有挑战性的数据集上的实验表明，CIPCMC优于现有的方法，在Caltech-3M数据集上实现了10.0%的准确率提高，在PBMC数据集上实现了16.6%的准确率提高。

{"title":"Cross-modal information propagation for contrastive multi-modal clustering","authors":"Tongji Chen , Guoliang Zou , Shizhe Hu, Yangdong Ye","doi":"10.1016/j.ipm.2025.104595","DOIUrl":"10.1016/j.ipm.2025.104595","url":null,"abstract":"<div><div>Multi-modal clustering aims to exploit relationships between different modalities to enhance clustering performance. However, existing methods face two main challenges. First, feature extraction fails to fully utilize the relationships between samples from different modalities, which is crucial for capturing global multi-modal information. Second, most clustering methods struggle to correct significantly erroneous assignments. To address these challenges, we propose Cross-modal Information Propagation for Contrastive Multi-modal Clustering (CIPCMC), a novel method driven by cross-modal information (CMI) and contrastive learning. We progressively obtain private CMI and integrate it into a unified CMI, which is then propagated to optimize the entire model. First, a cross-attention mechanism introduces CMI for each modality, enabling the model to focus on relationships between different modalities. This allows the model to uncover semantic associations and effectively exploit the complementary nature of multi-modal data. Next, we fuse modality-specific representations to derive a unified CMI representation, which helps each modality correct erroneous assignments, leading to high-confidence clustering. The end-to-end training of CIPCMC ensures module synergy, improving performance and generalization. Experiments on challenging datasets show that CIPCMC outperforms existing methods, achieving accuracy improvements of 10.0% on the Caltech-3M dataset and 16.6% on the PBMC dataset.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104595"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927588","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

Reevaluating zero-shot information extraction: Sampling bias, prompting transferability and sensitivity in large language models 重新评估零采样信息提取：大型语言模型中的抽样偏差、提示可转移性和敏感性

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-09 DOI: 10.1016/j.ipm.2026.104611

Ke Huang , Chenghao Xiao , Yao Xiao , Ming Cai , Noura Al Moubayed

Large Language Models (LLMs) have advanced zero-shot Information Extraction (IE), particularly in Sentence-level Relation Extraction (SentRE), through in-context learning and instruction tuning. However, the current evaluation of LLMs’ zero-shot ability on IE tasks remains fragile and unreliable. In this work, we provide a systematic examination of the fragility underlying current evaluation practices across three interrelated levels. At the data level, we demonstrate that the commonly adopted random sampling strategy introduces significant biases in class-imbalanced datasets, whereas balanced sampling provides more stable and faithful assessments of LLMs performance. At the task level, we reveal that three domain prompt frameworks on SentRE transfer inconsistently to Document-level Relation Extraction (DocRE) and Named Entity Recognition (NER), showing partial effectiveness on NER but notable limitations on DocRE due to long contexts and complex entity structures. At the method level, through extensive experiments on three IE tasks and seven datasets, we conduct the first comprehensive comparison of five general prompt frameworks, including Chain-of-Thought, Self-Improvement, and Self-Debate, showing that prompt effectiveness is highly task-dependent, with no single strategy dominating across tasks. For each task, the CoT prompt framework achieves the best performance on SentRE, the Vanilla prompt framework performs best on DocRE, and the Self-Consistency prompt framework excels on NER. These insights challenge current landscape of information extraction, providing guidelines for robust evaluation and prompt designs.

大型语言模型（llm）具有先进的零采样信息提取（IE），特别是在句子级关系提取（SentRE）中，通过上下文学习和指令调整。然而，目前对法学硕士在IE任务上的零射击能力的评估仍然是脆弱和不可靠的。在这项工作中，我们在三个相互关联的层面上对当前评估实践的脆弱性进行了系统检查。在数据层面上，我们证明了通常采用的随机抽样策略在类不平衡数据集中引入了显著的偏差，而平衡抽样提供了更稳定和可靠的llm性能评估。在任务层面上，我们揭示了SentRE上的三个领域提示框架在文档级关系提取（DocRE）和命名实体识别（NER）上的转换不一致，在NER上显示出部分有效性，但由于长上下文和复杂的实体结构，在DocRE上显示出明显的局限性。在方法层面，通过对三个IE任务和七个数据集的广泛实验，我们首次对五种一般提示框架（包括思维链、自我完善和自我辩论）进行了全面比较，表明提示有效性高度依赖于任务，没有单一策略在任务中占主导地位。对于每个任务，CoT提示框架在SentRE上表现最好，Vanilla提示框架在DocRE上表现最好，自一致性提示框架在NER上表现最好。这些见解挑战了当前的信息提取领域，为稳健的评估和快速的设计提供了指导方针。

{"title":"Reevaluating zero-shot information extraction: Sampling bias, prompting transferability and sensitivity in large language models","authors":"Ke Huang , Chenghao Xiao , Yao Xiao , Ming Cai , Noura Al Moubayed","doi":"10.1016/j.ipm.2026.104611","DOIUrl":"10.1016/j.ipm.2026.104611","url":null,"abstract":"<div><div>Large Language Models (LLMs) have advanced zero-shot Information Extraction (IE), particularly in Sentence-level Relation Extraction (SentRE), through in-context learning and instruction tuning. However, the current evaluation of LLMs’ zero-shot ability on IE tasks remains fragile and unreliable. In this work, we provide a systematic examination of the fragility underlying current evaluation practices across three interrelated levels. At the data level, we demonstrate that the commonly adopted random sampling strategy introduces significant biases in class-imbalanced datasets, whereas balanced sampling provides more stable and faithful assessments of LLMs performance. At the task level, we reveal that three domain prompt frameworks on SentRE transfer inconsistently to Document-level Relation Extraction (DocRE) and Named Entity Recognition (NER), showing partial effectiveness on NER but notable limitations on DocRE due to long contexts and complex entity structures. At the method level, through extensive experiments on three IE tasks and seven datasets, we conduct the first comprehensive comparison of five general prompt frameworks, including Chain-of-Thought, Self-Improvement, and Self-Debate, showing that prompt effectiveness is highly task-dependent, with no single strategy dominating across tasks. For each task, the CoT prompt framework achieves the best performance on SentRE, the Vanilla prompt framework performs best on DocRE, and the Self-Consistency prompt framework excels on NER. These insights challenge current landscape of information extraction, providing guidelines for robust evaluation and prompt designs.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104611"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927678","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 clustering based on the association of graph structure and feature distribution 基于图结构与特征分布关联的多视图聚类

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-02 DOI: 10.1016/j.ipm.2025.104586

Chenhui Shi , Yongjie Xin , Haifeng Yang , Jianghui Cai , Jie Wang , Lichan Zhou , Yanting He , Fuxing Cui , Xujun Zhao , Yaling Xun

Graph-based multi-view clustering methods have gained considerable attention in recent years. However, most existing techniques ignore the association of graph and feature distributions between different views. In addition, noise and redundant information in data will leads to an inability to accurately learn consistent distributions among multiple views. To overcome these issues, this study proposes a framework termed “multi-view clustering based on the association of graph structure and feature distribution” (MLGF). Specifically, we provide collaborative training based on a similar distribution comparison mechanism that unifies the graph structures and feature distributions of different views, to construct multiple high-quality similarity matrices. Noisy information is effectively eliminated from the raw data by embedding graph spectral decomposition and automatic weighting methods into the graph encoder to learn clean, low-dimensional embedded representations of the data. Finally, multiple similarity matrices are fused in a locally weighted manner to obtain consistent similarity matrices. Experiments on five benchmark datasets demonstrated the superiority of our method, achieving 100%, 97.28% on COIL-20 and Handwritten datasets. This is attributed to the effective joint optimization of graph structure and feature distribution, which is validated by its outstanding performance across diverse datasets. The code will be available at https://github.com/shichenhui/MLGF.

基于图的多视图聚类方法近年来得到了广泛的关注。然而，大多数现有的技术忽略了不同视图之间的图和特征分布的关联。此外，数据中的噪声和冗余信息将导致无法准确地学习多个视图之间的一致分布。为了克服这些问题，本研究提出了一种基于图结构和特征分布关联的多视图聚类框架（MLGF）。具体来说，我们提供了基于相似分布比较机制的协同训练，该机制统一了不同视图的图结构和特征分布，以构建多个高质量的相似矩阵。通过将图谱分解和自动加权方法嵌入到图编码器中，以学习数据的干净、低维嵌入表示，有效地消除了原始数据中的噪声信息。最后，对多个相似矩阵进行局部加权融合，得到一致性相似矩阵。在5个基准数据集上的实验证明了该方法的优越性，在COIL-20和手写数据集上的准确率分别为100%、97.28%。这归功于图结构和特征分布的有效联合优化，其在不同数据集上的出色性能验证了这一点。代码可在https://github.com/shichenhui/MLGF上获得。

{"title":"Multi-view clustering based on the association of graph structure and feature distribution","authors":"Chenhui Shi , Yongjie Xin , Haifeng Yang , Jianghui Cai , Jie Wang , Lichan Zhou , Yanting He , Fuxing Cui , Xujun Zhao , Yaling Xun","doi":"10.1016/j.ipm.2025.104586","DOIUrl":"10.1016/j.ipm.2025.104586","url":null,"abstract":"<div><div>Graph-based multi-view clustering methods have gained considerable attention in recent years. However, most existing techniques ignore the association of graph and feature distributions between different views. In addition, noise and redundant information in data will leads to an inability to accurately learn consistent distributions among multiple views. To overcome these issues, this study proposes a framework termed “multi-view clustering based on the association of graph structure and feature distribution” (MLGF). Specifically, we provide collaborative training based on a similar distribution comparison mechanism that unifies the graph structures and feature distributions of different views, to construct multiple high-quality similarity matrices. Noisy information is effectively eliminated from the raw data by embedding graph spectral decomposition and automatic weighting methods into the graph encoder to learn clean, low-dimensional embedded representations of the data. Finally, multiple similarity matrices are fused in a locally weighted manner to obtain consistent similarity matrices. Experiments on five benchmark datasets demonstrated the superiority of our method, achieving 100%, 97.28% on COIL-20 and Handwritten datasets. This is attributed to the effective joint optimization of graph structure and feature distribution, which is validated by its outstanding performance across diverse datasets. The code will be available at <span><span>https://github.com/shichenhui/MLGF</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104586"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145886174","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 multi-criteria sorting method for preference maps based on Nash-Stackelberg game 基于Nash-Stackelberg博弈的偏好图多准则排序方法

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-02 DOI: 10.1016/j.ipm.2025.104587

Xinru Han , Yukun Bao , Jianming Zhan , Yufeng Shen

Existing multi-criteria sorting methods predominantly rely on preset classification thresholds or fixed numbers of alternatives for classification, exhibiting strong subjectivity and overlooking potential consensus correlations between classifications. In group decision-making (GDM), the consensus feedback mechanism drives the consensus reaching process (CRP) and gives rise to the problem of adjustment amount allocation among decision-makers (DMs). However, existing studies over-rely on consensus thresholds and neglect differences in DMs’ adjustment capabilities and sequences, which significantly reduces the applicability and accuracy of the methods. To address the above issues, this study proposes a novel group consensus method (NS-FPR-PM) integrating the Nash-Stackelberg game and preference maps within the framework of fuzzy preference relations (FPRs). Specifically, class probability thresholds are objectively derived through an optimization model; the classification results are then converted into preference maps based on these class probability thresholds to explore the inherent consensus relations, thereby eliminating reliance on consensus thresholds. The Nash-Stackelberg game model can characterize the differences in bargaining power among DMs, and an asynchronous adjustment mechanism is designed accordingly to achieve fair allocation of adjustment amount. Finally, we provide an example to illustrate the proposed method, the experimental results and analysis demonstrate that the method exhibits significant advantages over similar methods in terms of consensus reaching efficiency and unit adjustment conversion rate.

现有的多标准分类方法主要依赖于预设的分类阈值或固定数量的备选分类，表现出很强的主观性，忽略了分类之间潜在的共识相关性。在群体决策（GDM）中，共识反馈机制推动了共识达成过程（CRP），并产生了决策者之间调整量分配的问题。然而，现有研究过度依赖共识阈值，忽视了dm调整能力和序列的差异，大大降低了方法的适用性和准确性。为了解决上述问题，本研究提出了一种新的群体共识方法（NS-FPR-PM），该方法将纳什- stackelberg博弈和模糊偏好关系（FPRs）框架下的偏好图相结合。具体而言，通过优化模型客观地推导出类概率阈值；然后将分类结果转换为基于这些类概率阈值的偏好图，以探索固有的共识关系，从而消除对共识阈值的依赖。Nash-Stackelberg博弈模型可以表征dm之间议价能力的差异，并据此设计异步调整机制，实现调整金额的公平分配。最后，给出了一个算例，实验结果和分析表明，该方法在共识达成效率和单位调整转化率方面比同类方法具有显著的优势。

{"title":"A multi-criteria sorting method for preference maps based on Nash-Stackelberg game","authors":"Xinru Han , Yukun Bao , Jianming Zhan , Yufeng Shen","doi":"10.1016/j.ipm.2025.104587","DOIUrl":"10.1016/j.ipm.2025.104587","url":null,"abstract":"<div><div>Existing multi-criteria sorting methods predominantly rely on preset classification thresholds or fixed numbers of alternatives for classification, exhibiting strong subjectivity and overlooking potential consensus correlations between classifications. In group decision-making (GDM), the consensus feedback mechanism drives the consensus reaching process (CRP) and gives rise to the problem of adjustment amount allocation among decision-makers (DMs). However, existing studies over-rely on consensus thresholds and neglect differences in DMs’ adjustment capabilities and sequences, which significantly reduces the applicability and accuracy of the methods. To address the above issues, this study proposes a novel group consensus method (NS-FPR-PM) integrating the Nash-Stackelberg game and preference maps within the framework of fuzzy preference relations (FPRs). Specifically, class probability thresholds are objectively derived through an optimization model; the classification results are then converted into preference maps based on these class probability thresholds to explore the inherent consensus relations, thereby eliminating reliance on consensus thresholds. The Nash-Stackelberg game model can characterize the differences in bargaining power among DMs, and an asynchronous adjustment mechanism is designed accordingly to achieve fair allocation of adjustment amount. Finally, we provide an example to illustrate the proposed method, the experimental results and analysis demonstrate that the method exhibits significant advantages over similar methods in terms of consensus reaching efficiency and unit adjustment conversion rate.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104587"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145886177","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 text-based emotional pattern discrepancy aware model for enhanced generalization in depression detection 基于文本的情绪模式差异感知模型在抑郁症检测中的增强泛化

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2025-12-31 DOI: 10.1016/j.ipm.2025.104575

Haibo Zhang , Zhenyu Liu , Yang Wu , Jiaqian Yuan , Gang Li , Zhijie Ding , Bin Hu

Text-based automated depression detection is one of the current hot topics. However, current research lacks the exploration of key verbal behaviors in depression detection scenarios, resulting in insufficient generalization performance of the models. To address this issue, we propose a depression detection method based on emotional pattern discrepancies, as the discrepancies are one of the fundamental features of depression as an affective disorder. Specifically, we propose an Emotional Pattern Discrepancy Aware Depression Detection Model (EPDAD). The EPDAD employs specially designed modules and loss functions to train the model. This approach enables the model to dynamically and comprehensively perceive the different emotional patterns reflected by depressed and healthy individuals in response to various emotional stimuli. As a result, it enhances the model’s ability to learn the essential features of depression. We evaluate the generalization performance of our model from a cross-dataset and cross-topic perspective using MODMA (52 samples) and MIDD (520 samples) datasets. In cross-topic generalization experiments, our method improves F1 score by 10.39% and 1.77% on MODMA and MIDD, respectively, in comparison to the state-of-the-art method. In cross-dataset generalization experiments, our method improves the F1 score by a maximum of 6.37%. We also compare our model with large language models, and the results indicate it is more effective for depression detection tasks. Our research contributes to the practical application of depression detection models. Our code is available at: https://github.com/hbZhzzz/EPDAD.

基于文本的抑郁症自动检测是当前研究的热点之一。然而，目前的研究缺乏对抑郁症检测场景中关键言语行为的探索，导致模型的泛化性能不足。为了解决这个问题，我们提出了一种基于情绪模式差异的抑郁症检测方法，因为差异是抑郁症作为一种情感障碍的基本特征之一。具体而言，我们提出了一个情绪模式差异感知抑郁检测模型（EPDAD）。EPDAD采用专门设计的模块和损失函数对模型进行训练。该方法使模型能够动态、全面地感知抑郁个体和健康个体对各种情绪刺激所反映的不同情绪模式。因此，它增强了模型学习抑郁症基本特征的能力。我们使用MODMA（52个样本）和MIDD（520个样本）数据集从跨数据集和跨主题的角度评估了我们的模型的泛化性能。在交叉主题泛化实验中，我们的方法在MODMA和MIDD上分别提高了10.39%和1.77%的F1分数。在跨数据集泛化实验中，我们的方法将F1分数提高了6.37%。我们还将我们的模型与大型语言模型进行了比较，结果表明它在抑郁检测任务中更有效。我们的研究有助于抑郁症检测模型的实际应用。我们的代码可在：https://github.com/hbZhzzz/EPDAD。

{"title":"A text-based emotional pattern discrepancy aware model for enhanced generalization in depression detection","authors":"Haibo Zhang , Zhenyu Liu , Yang Wu , Jiaqian Yuan , Gang Li , Zhijie Ding , Bin Hu","doi":"10.1016/j.ipm.2025.104575","DOIUrl":"10.1016/j.ipm.2025.104575","url":null,"abstract":"<div><div>Text-based automated depression detection is one of the current hot topics. However, current research lacks the exploration of key verbal behaviors in depression detection scenarios, resulting in insufficient generalization performance of the models. To address this issue, we propose a depression detection method based on emotional pattern discrepancies, as the discrepancies are one of the fundamental features of depression as an affective disorder. Specifically, we propose an Emotional Pattern Discrepancy Aware Depression Detection Model (EPDAD). The EPDAD employs specially designed modules and loss functions to train the model. This approach enables the model to dynamically and comprehensively perceive the different emotional patterns reflected by depressed and healthy individuals in response to various emotional stimuli. As a result, it enhances the model’s ability to learn the essential features of depression. We evaluate the generalization performance of our model from a cross-dataset and cross-topic perspective using MODMA (52 samples) and MIDD (520 samples) datasets. In cross-topic generalization experiments, our method improves F1 score by 10.39% and 1.77% on MODMA and MIDD, respectively, in comparison to the state-of-the-art method. In cross-dataset generalization experiments, our method improves the F1 score by a maximum of 6.37%. We also compare our model with large language models, and the results indicate it is more effective for depression detection tasks. Our research contributes to the practical application of depression detection models. Our code is available at: <span><span>https://github.com/hbZhzzz/EPDAD</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104575"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145886271","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

Validating generative agent-Based modeling in social media simulations through the lens of the friendship paradox 通过友谊悖论的镜头验证社交媒体模拟中基于生成代理的建模

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-23 DOI: 10.1016/j.ipm.2026.104636

Gian Marco Orlando , Valerio La Gatta , Diego Russo , Vincenzo Moscato

Generative Agent-Based Modeling (GABM) is an emerging simulation paradigm that integrates the reasoning capabilities of Large Language Models (LLMs) with traditional Agent-Based Modeling to replicate complex social behaviors, including user interactions on social media platforms. While GABM has been employed to study localized phenomena on social media, such as opinion formation and information propagation, its capacity to capture global network-level phenomena remains underexplored. In this paper, we address this gap by investigating whether GABM-based social media simulations exhibit the Friendship Paradox (FP) – a counterintuitive network phenomenon where individuals, on average, have fewer friends than their friends. We design a GABM-based framework for social media simulation, featuring generative agents that emulate real users by incorporating distinct personalities, interests, and behaviors. Leveraging three real-world Twitter datasets centered on the US 2020 Election, UK Brexit, and the QAnon conspiracy, we demonstrate that the FP and its generalized forms emerge in GABM-based simulations. Consistent with real-world social media, we observe a hierarchical structure where generative agents preferentially connect with others exhibiting superior attributes, such as greater activity or influence, without being instructed with any behavioral rules. Furthermore, our analysis reveals that infrequent connections with highly connected agents primarily drive the Friendship Paradox, mirroring established patterns in real-world networks. Overall, our findings validate the ability of GABM to replicate global social media phenomena, highlighting its potential as a robust framework for modeling and analyzing complex social behaviors at scale.

基于生成代理的建模（GABM）是一种新兴的仿真范式，它将大型语言模型（llm）的推理能力与传统的基于代理的建模相结合，以复制复杂的社会行为，包括社交媒体平台上的用户交互。虽然GABM已被用于研究社交媒体上的局部现象，如意见形成和信息传播，但其捕捉全球网络层面现象的能力仍未得到充分探索。在本文中，我们通过调查基于gabm的社交媒体模拟是否表现出友谊悖论（FP）来解决这一差距。FP是一种反直觉的网络现象，平均而言，个人的朋友比他们的朋友少。我们设计了一个基于gabm的社交媒体模拟框架，其特点是生成代理通过结合不同的个性、兴趣和行为来模拟真实用户。利用以2020年美国大选、英国脱欧和QAnon阴谋为中心的三个真实世界的Twitter数据集，我们证明了FP及其广义形式出现在基于gabm的模拟中。与现实世界的社交媒体一致，我们观察到一种层次结构，在这种结构中，生成智能体优先与表现出更优越属性的其他智能体连接，比如更活跃或更有影响力，而不受任何行为规则的指导。此外，我们的分析表明，与高度连接的代理之间的不频繁连接主要驱动了友谊悖论，反映了现实世界网络中的既定模式。总的来说，我们的研究结果验证了GABM复制全球社交媒体现象的能力，突出了它作为大规模建模和分析复杂社会行为的强大框架的潜力。

{"title":"Validating generative agent-Based modeling in social media simulations through the lens of the friendship paradox","authors":"Gian Marco Orlando , Valerio La Gatta , Diego Russo , Vincenzo Moscato","doi":"10.1016/j.ipm.2026.104636","DOIUrl":"10.1016/j.ipm.2026.104636","url":null,"abstract":"<div><div>Generative Agent-Based Modeling (GABM) is an emerging simulation paradigm that integrates the reasoning capabilities of Large Language Models (LLMs) with traditional Agent-Based Modeling to replicate complex social behaviors, including user interactions on social media platforms. While GABM has been employed to study localized phenomena on social media, such as opinion formation and information propagation, its capacity to capture global network-level phenomena remains underexplored. In this paper, we address this gap by investigating whether GABM-based social media simulations exhibit the Friendship Paradox (FP) – a counterintuitive network phenomenon where individuals, on average, have fewer friends than their friends. We design a GABM-based framework for social media simulation, featuring <em>generative agents</em> that emulate real users by incorporating distinct personalities, interests, and behaviors. Leveraging three real-world Twitter datasets centered on the US 2020 Election, UK Brexit, and the QAnon conspiracy, we demonstrate that the FP and its generalized forms emerge in GABM-based simulations. Consistent with real-world social media, we observe a hierarchical structure where <em>generative agents</em> preferentially connect with others exhibiting superior attributes, such as greater activity or influence, without being instructed with any behavioral rules. Furthermore, our analysis reveals that infrequent connections with highly connected agents primarily drive the Friendship Paradox, mirroring established patterns in real-world networks. Overall, our findings validate the ability of GABM to replicate global social media phenomena, highlighting its potential as a robust framework for modeling and analyzing complex social behaviors at scale.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104636"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146023310","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

Dual-stream spatiotemporal graph convolutional networks for EEG-based human emotion recognition 基于脑电图的人类情感识别的双流时空图卷积网络

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-08 DOI: 10.1016/j.ipm.2025.104597

Jiaying Ren , Fengming Han , Yadong Xu

Deep learning has advanced EEG-based human emotion recognition, yet most existing approaches rely on either temporal or spectral features and insufficiently model the fine-grained spatiotemporal structure of neural activity. To address these challenges, this paper develops a dual-stream spatiotemporal graph convolutional network (DSSGCN) for human emotion recognition. In the time domain, a multi-scale modern temporal convolutional network (MS-MTCN) is designed to capture rich temporal information across diverse receptive fields and model long-range temporal dependencies. In the frequency domain, a fully-connected multi-scale graph attention network (FM-GAT) is introduced to learn complex inter-channel relationships and spatial dependencies from the spectral representation of EEG signals. Furthermore, a cross-domain feature fusion module (CFFM) is employed to integrate the complementary information from both temporal and spectral branches, followed by an adaptive ensemble classifier (AEC) to enhance recognition robustness. Finally, an improved online knowledge distillation (IOKD) algorithm is devised to enhance the model’s robustness and generalization. Evaluated on two public dataset and a self-collected music-emotion dataset, DSSGCN achieves 93.98%, 85.00%, and 99.20% accuracy, consistently surpassing eleven state-of-the-art methods and validating its effectiveness for decoding affective states from EEG signals.

深度学习促进了基于脑电图的人类情感识别，但大多数现有方法依赖于时间或频谱特征，无法充分模拟神经活动的细粒度时空结构。为了解决这些挑战，本文开发了一种用于人类情感识别的双流时空图卷积网络（DSSGCN）。在时域上，设计了一种多尺度现代时间卷积网络（MS-MTCN），以捕获跨不同感受野的丰富时间信息，并对长时间依赖关系进行建模。在频域，引入全连接多尺度图注意网络（FM-GAT），从脑电信号的频谱表征中学习复杂的通道间关系和空间依赖关系。在此基础上，采用跨域特征融合模块（CFFM）对时间分支和光谱分支的互补信息进行融合，并采用自适应集成分类器（AEC）增强识别的鲁棒性。最后，设计了一种改进的在线知识蒸馏（IOKD）算法来增强模型的鲁棒性和泛化性。在两个公共数据集和一个自我收集的音乐情感数据集上进行评估，DSSGCN达到了93.98%，85.00%和99.20%的准确率，持续超过了11种最先进的方法，并验证了其从EEG信号中解码情感状态的有效性。

{"title":"Dual-stream spatiotemporal graph convolutional networks for EEG-based human emotion recognition","authors":"Jiaying Ren , Fengming Han , Yadong Xu","doi":"10.1016/j.ipm.2025.104597","DOIUrl":"10.1016/j.ipm.2025.104597","url":null,"abstract":"<div><div>Deep learning has advanced EEG-based human emotion recognition, yet most existing approaches rely on either temporal or spectral features and insufficiently model the fine-grained spatiotemporal structure of neural activity. To address these challenges, this paper develops a dual-stream spatiotemporal graph convolutional network (DSSGCN) for human emotion recognition. In the time domain, a multi-scale modern temporal convolutional network (MS-MTCN) is designed to capture rich temporal information across diverse receptive fields and model long-range temporal dependencies. In the frequency domain, a fully-connected multi-scale graph attention network (FM-GAT) is introduced to learn complex inter-channel relationships and spatial dependencies from the spectral representation of EEG signals. Furthermore, a cross-domain feature fusion module (CFFM) is employed to integrate the complementary information from both temporal and spectral branches, followed by an adaptive ensemble classifier (AEC) to enhance recognition robustness. Finally, an improved online knowledge distillation (IOKD) algorithm is devised to enhance the model’s robustness and generalization. Evaluated on two public dataset and a self-collected music-emotion dataset, DSSGCN achieves 93.98%, 85.00%, and 99.20% accuracy, consistently surpassing eleven state-of-the-art methods and validating its effectiveness for decoding affective states from EEG signals.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104597"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927585","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

End-to-end scheduling for carrier-based aircraft sortie operations using deep reinforcement learning 基于深度强化学习的舰载机出动作战端到端调度

IF 6.9 1区管理学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Information Processing & Management

Pub Date : 2026-06-01 Epub Date: 2026-01-05 DOI: 10.1016/j.ipm.2025.104590

Changjiu Li , Wei Han , Yong Zhang , Xinwei Wang , Xichao Su

Efficient scheduling of carrier-based aircraft sorties is essential for enhancing the effectiveness of aircraft carriers. The key research challenges stem from the limitations of traditional algorithms, which struggle with this complex scheduling problem due to their high computational complexity, poor adaptability to dynamic events, and a tendency to converge to local optima, rendering them unsuitable for meeting real-time operational demands. To tackle these challenges, we propose an end-to-end deep reinforcement learning scheduling framework that leverages a multi-head attention mechanism to extract features from a heterogeneous graph of the scheduling environment. Using the proximal policy optimization-clip algorithm, the framework enables iterative interaction with a simulation environment to train the scheduling agent. Our experimental findings quantitatively demonstrate the superiority of the proposed framework: the agent outperforms traditional combined rules by over 5% and metaheuristic algorithms by approximately 1%, while achieving an average decision-making time of just 0.7 seconds. The model also demonstrates strong robustness, maintaining a minimal optimality gap even under a 30% reduction in resources. This research provides commanders with a more efficient decision support tool, thereby improving their battlefield response capabilities.

有效的舰载机出动调度是提高航母战斗力的关键。关键的研究挑战源于传统算法的局限性，传统算法由于计算量大、对动态事件的适应性差、倾向于收敛于局部最优而无法满足实时操作需求，难以解决复杂的调度问题。为了解决这些挑战，我们提出了一个端到端的深度强化学习调度框架，该框架利用多头注意机制从调度环境的异构图中提取特征。该框架采用近端策略优化-剪辑算法，实现了与仿真环境的迭代交互，以训练调度代理。我们的实验结果定量地证明了所提出框架的优越性：智能体比传统的组合规则高出5%以上，比元启发式算法高出约1%，而平均决策时间仅为0.7秒。该模型还显示出很强的鲁棒性，即使在资源减少30%的情况下，也能保持最小的最优性差距。本研究为指挥官提供了更有效的决策支持工具，从而提高了他们的战场响应能力。

{"title":"End-to-end scheduling for carrier-based aircraft sortie operations using deep reinforcement learning","authors":"Changjiu Li , Wei Han , Yong Zhang , Xinwei Wang , Xichao Su","doi":"10.1016/j.ipm.2025.104590","DOIUrl":"10.1016/j.ipm.2025.104590","url":null,"abstract":"<div><div>Efficient scheduling of carrier-based aircraft sorties is essential for enhancing the effectiveness of aircraft carriers. The key research challenges stem from the limitations of traditional algorithms, which struggle with this complex scheduling problem due to their high computational complexity, poor adaptability to dynamic events, and a tendency to converge to local optima, rendering them unsuitable for meeting real-time operational demands. To tackle these challenges, we propose an end-to-end deep reinforcement learning scheduling framework that leverages a multi-head attention mechanism to extract features from a heterogeneous graph of the scheduling environment. Using the proximal policy optimization-clip algorithm, the framework enables iterative interaction with a simulation environment to train the scheduling agent. Our experimental findings quantitatively demonstrate the superiority of the proposed framework: the agent outperforms traditional combined rules by over 5% and metaheuristic algorithms by approximately 1%, while achieving an average decision-making time of just 0.7 seconds. The model also demonstrates strong robustness, maintaining a minimal optimality gap even under a 30% reduction in resources. This research provides commanders with a more efficient decision support tool, thereby improving their battlefield response capabilities.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 4","pages":"Article 104590"},"PeriodicalIF":6.9,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927586","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