Information Processing & Management最新文献

{"title":"DF-NeRf: A geometrically consistent 3D reconstruction method for blurry images with decoupled and spatial fusion","authors":"Xiaohui Wang ,&nbsp;Zhenyu Yin ,&nbsp;Feiqing Zhang","doi":"10.1016/j.ipm.2026.104669","DOIUrl":"10.1016/j.ipm.2026.104669","url":null,"abstract":"<div><div>Reliable 3D scene understanding is fundamental to intelligent information systems, but motion blur in visual data often disrupts geometric consistency, making the reconstruction results unreliable. This paper proposes a novel blurred image NeRF framework named DF-NeRF, incorporating three core innovations. Firstly, the Multi-Scale Position Feature Fusion (MSPF) structure performs multi-scale feature fusion within the blurred kernel space to precisely estimate ray displacement ranges, thereby preserving multi-view geometric consistency. Secondly, the Multi-layer Independent Weight and Shifting Network (MIWS) is designed to achieve explicit decoupling between geometry and appearance. Within MIWS, the Dual-head Network (DHNet) employs a separate optimisation strategy, independently learning ray offsets (geometric attributes) and rendering weights (appearance attributes), effectively suppressing geometric distortions caused by motion blur. Meanwhile, the Prior-weighted Network (PWNet) employs prior-guided adaptive weighting to fuse different blur kernels, ensuring high-frequency detail geometric continuity. Finally, an edge-aware loss function is introduced to constrain gradient discrepancies in depth discontinuities, explicitly guaranteeing contour structure consistency and detail fidelity. We evaluated on public benchmarks, including 5 synthetic scenes and 10 real scenes, covering both indoor and outdoor environments. Experimental results demonstrate that DF-NeRF outperforms SOTA on PSNR, SSIM, and LPIPS metrics, with the most notable improvement in LPIPS. In synthetic scenes, DF-NeRF achieves an average LPIPS improvement of 14.32% over SOTA. In real-world scenes, DF-NeRF achieves an average LPIPS improvement of 19.79% over SOTA. DF-NeRF significantly enhances 3D reconstruction quality and robustness under blurred images, making it suitable for applications such as autonomous driving and robot vision. Code is available at: <span><span>https://github.com/luckhui0505/DF-NeRF</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104669"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190663","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}

{"title":"Dynamic loss balancing for long-tailed multi-label problem in document-level relation extraction","authors":"Weiwei Zhang ,&nbsp;Meiling Xu ,&nbsp;Te Xu","doi":"10.1016/j.ipm.2026.104656","DOIUrl":"10.1016/j.ipm.2026.104656","url":null,"abstract":"<div><div>Documents contain multiple entity pairs, with specific pairs associated with multiple relations. The long-tailed distribution of relations further exacerbates the complexity of the document-level relation extraction (DocRE) task. In this study, we introduce a novel DocRE framework called HeterDLB. The HeterDLB constructs mention-level heterogeneous graphs by employing two types of meta-paths to capture the rich structural and contextual information within the document. To address the long-tailed distribution in multi-label DocRE, we propose an adaptive confidence threshold strategy that effectively distinguishes between positive and negative classes. To mitigate the imbalance between positive and negative classes during the training process, we first define the gradient update process of HeterDLB as a multi-objective optimization problem that involves adjusting the penalty strength among various loss functions through dynamic weights. We evaluate the HeterDLB on four benchmark datasets: DocRED, Re-DocRED, and two biomedical datasets, CDR and GDA. HeterDLB achieves <em>F</em>₁ scores of 65.96% and 82.20% on DocRED and Re-DocRED, and 70.95% and 85.18% on CDR and GDA, respectively. Compared to the recent models, HeterDLB achieves average <em>F</em>₁ score improvements of 5.43%, 15.48%, 4.07%, and 7.85% on the respective datasets when using the BERT_{<em>base</em>} model. Extensive experiments from multiple perspectives further validate the effectiveness of our framework.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104656"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190668","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}

{"title":"CLER: A benchmark for Chinese litigation evidence reasoning","authors":"Sheng Bi ,&nbsp;Kai Yu ,&nbsp;Jiaqi Li ,&nbsp;Xiaoyan Wang ,&nbsp;Fuhui Sun ,&nbsp;Zeyi Miao ,&nbsp;Jing Wang ,&nbsp;Xin Geng ,&nbsp;Lusheng Wang","doi":"10.1016/j.ipm.2026.104667","DOIUrl":"10.1016/j.ipm.2026.104667","url":null,"abstract":"<div><div>Evidence prediction is the cornerstone of litigation, which requires complex legal reasoning to bridge the gap between facts and the parties’ assertions. However, most existing large language model (LLM) benchmarks do not offer a systematic assessment for evidence planning. Therefore, we introduce CLER, a large-scale Chinese Litigation Evidence Reasoning benchmark. CLER is built through a four-stage pipeline from millions of real civil and criminal judgments. The pipeline includes (1) semi-structured data collection, (2) LLM-based evidence extraction, (3) purpose refinement through decomposition into legal elements, and (4) validation through a tri-party adversarial review method that imitates courtroom debate. To this end, we obtain over 100,000 samples and conduct extensive experiments on 23 baselines, including open-weight, commercial, domain-specific LLMs and retrieval-augmented generation (RAG) systems. The findings indicate that the task poses a major challenge for current LLMs. RAG strategies, particularly a hybrid method retrieving similar cases, achieve the best performance with an F1 score of 54.4%, surpassing state-of-the-art proprietary models and improving the base model by at least 22%. Our analysis reveals several key insights. There is a significant performance gap with an average difference F1 of 16.2% exists between element checking in civil cases and narrative completeness in criminal cases. There is an inverted U-shaped correlation between case length and model performance, as insufficient or overloaded information leads to abandonment of precise reasoning, causing an average drop of approximately 10% in F1.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104667"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190673","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}

{"title":"Positive and negative neighbor dual-flexible nonnegative representation method for image classification","authors":"Ziqi Li ,&nbsp;Zhilin Chen ,&nbsp;Tingting Guo ,&nbsp;Yonghong Zhang ,&nbsp;Xiaoning Song ,&nbsp;Tianyang Xu","doi":"10.1016/j.ipm.2026.104645","DOIUrl":"10.1016/j.ipm.2026.104645","url":null,"abstract":"<div><div>Nonnegative representation-based classification (NRC) enhances the discriminability of samples to some extent through nonnegativity constraints. However, this approach often fails to fully exploit local feature structures and class-discriminative information, which limits its overall classification performance. To overcome this limitation, we introduce the positive and negative neighbor dual-flexible nonnegative representation (PN²DFNR) classifier. This method integrates both the positive and negative neighbor sets of the query sample to effectively capture location information. Specifically, the positive neighbor set enforces spatial consistency, ensuring that the estimated representation of the query sample remains well-aligned with its neighboring samples. In contrast, the negative neighbor set introduces inverse constraints to suppress inter-class interference. Furthermore, a flexibility factor is incorporated to formulate a weighted flexible constraint strategy, which enhances the representational capacity of the correct class while adaptively attenuating the contribution of incorrect classes. To evaluate PN²DFNR’s performance, extensive experiments are conducted on facial recognition, handwritten digit classification, and large-scale datasets with diverse feature characteristics. The results demonstrate that PN²DFNR achieves superior classification performance, with the maximum improvement reaching approximately 2%. The model code will be available on the author’s homepage (<span><span>https://github.com/li-zi-qi/PN2DFNR</span><svg><path></path></svg></span>).</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104645"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146045206","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}

{"title":"A unified spatial-spectral pyramid mamba for hyperspectral image classification","authors":"Pei Du,&nbsp;Yu Huo,&nbsp;Chenhao Wang,&nbsp;Min Zhang,&nbsp;Hai Wang","doi":"10.1016/j.ipm.2026.104664","DOIUrl":"10.1016/j.ipm.2026.104664","url":null,"abstract":"<div><div>Hyperspectral image (HSI) classification is vital in remote sensing for land cover mapping, environmental surveillance, and precision agriculture. However, achieving high accuracy with minimal samples is a major challenge. Mamba’s superior long-range modeling and linear complexity offer a promising solution. Furthermore, leveraging the 3D structure of HSIs, effective multiscale spatial-spectral feature extraction is key to improving performance. Towards this end, we propose the Pyramid-based Spatial-Spectral Mamba (PS²Mamba). It has a dual-branch architecture within a unified framework with three key modules. In the spectral branch, the Pyramid-based Spectral Mamba (PSpeM) incrementally captures hierarchical spectral representations through spatial neighborhood weighting and a bidirectional scanning strategy to enhance spectral discrimination. The spatial branch’s Pyramid-based Spatial Mamba (PSpaM) captures multi-scale, multi-directional spatial features, incorporating an innovative eight-direction scanning mechanism to encode both fine and coarse patterns. In addition, a Spatial-Spectral Interaction Fusion (SSIF) module enables efficient cross-domain interaction for seamless fusion. Experimental results demonstrate that the proposed method achieves overall accuracy (OA) scores of 90.90%, 95.62%, 93.09%, 95.78% and 91.83% on five benchmark HSI datasets, under a limited training regime with only 10 labeled samples per class. Notably, these results represent significant improvements over state-of-the-art algorithms, with relative gains of 1.20%, 1.11%, 2.61%, 1.14% and 1.87% respectively. The code will be available at: <span><span>https://github.com/Duppy0509/PS2Mamba</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104664"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190664","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}

{"title":"Game-theoretic approach on strategizing cooperation model in community group-buying","authors":"Lingli Shu ,&nbsp;Yongyi Zhou ,&nbsp;Yong Qin","doi":"10.1016/j.ipm.2026.104680","DOIUrl":"10.1016/j.ipm.2026.104680","url":null,"abstract":"<div><div>Sound operational strategies and strong partnerships derive the development of community group-buying platform. Although the existing literature has concerned this unique platform, it largely neglects the design of its cooperation model. To fill this gap, this study proposes a cooperation framework based on a Stackelberg game approach in the context of community group-buying. Our model results demonstrate that collaborating with the local dealer through resale model enhances the platform’s profitability. Meanwhile, adopting a proportional commission model with the community leader can create a win-win-win situation involving the platform, the local dealer, and the community leader. Furthermore, a proportional commission structure exhibits superior incentive effects compared to a fixed commission structure in motivating community leader’s service efforts, especially when the local dealer operates in the resale model. Notably, our findings reveal that certain key parameters significantly influence the behavior of the platform’s partners, encouraging the local dealer and the community leader to adopt higher prices and exert greater effort in community group-buying service, while simultaneously providing opportunity for free-riding behavior. Experimental results confirm that the game-theoretic approach enables the platform to identify the optimal cooperation model, thereby achieving profit maximization.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104680"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190666","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}

{"title":"Enterprise efficiency analysis based on explainable artificial intelligence: From predictive algorithms to mechanisms","authors":"Hailin Li ,&nbsp;Hufeng Li ,&nbsp;Wenkai Shi ,&nbsp;Yenchun Jim Wu","doi":"10.1016/j.ipm.2026.104666","DOIUrl":"10.1016/j.ipm.2026.104666","url":null,"abstract":"<div><div>Enhancing corporate efficiency is a critical challenge in management research due to the complexity inherent in multi-stage processes. This study analyzes data from 29,602 corporate samples spanning 2009–2023, using super-efficiency Data Envelopment Analysis to decompose corporate efficiency into innovation and commercialization stages. Our analysis identified four distinct efficiency clusters, with only 27.667% of firms achieving high performance in both stages. We developed predictive models using nine machine learning algorithms, XGBoost outperformed the others, predicting corporate efficiency clusters with an accuracy of 72.520%. SHAP analysis revealed that firm age, financing constraints, and largest shareholder equity are the most critical predictive factors. Bayesian network analysis uncovered two key pathways influencing corporate innovation efficiency: “largest shareholder ownership impacting employee size impacting enterprise efficiency´´ and “debt-to-asset ratio impacting employee size impacting enterprise efficiency´´. These findings demonstrate that changes in individual variables can trigger cascading effects through network transmission, ultimately affecting overall corporate efficiency. Consequently, governments and enterprises should adopt differentiated strategies based on actual efficiency performance to facilitate efficiency improvements.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104666"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190672","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}

{"title":"Pontis: A decentralized framework for unifying remote attestation and enabling interoperability between heterogeneous TEEs","authors":"Jun Li ,&nbsp;Hong Lei ,&nbsp;Xinman Luo ,&nbsp;Lei Zhang ,&nbsp;Pengxu Shen ,&nbsp;Tianyu Huang ,&nbsp;Jieren Cheng","doi":"10.1016/j.ipm.2026.104652","DOIUrl":"10.1016/j.ipm.2026.104652","url":null,"abstract":"<div><div>In modern decentralized information systems, establishing verifiable trust in remote computing environments has emerged as a critical challenge for secure cross-domain collaboration. Hardware-based Trusted Execution Environments (TEEs) such as Intel SGX and ARM TrustZone offer a promising foundation for addressing this challenge through cryptographically verifiable execution guarantees, but their incompatible Remote Attestation (RA) mechanisms create fundamental barriers to cross-platform trust establishment. Current solutions either focus on single-vendor ecosystems or introduce prohibitive architectural complexity, failing to address the critical need for lightweight, decentralized interoperability. This paper presents Pontis, a decentralized blockchain-based framework that solves unified RA and cross-platform communication challenges for heterogeneous TEEs through three key innovations: (1) a distributed off-chain Coordinator that normalizes vendor-specific attestation protocols, combined with blockchain-anchored decentralized identifiers that provide immutable distributed identities for TEE instances; (2) blockchain-based smart contracts implementing Registration, Attestation, and Management functions for immutable trust status propagation; and (3) secure cross-TEE communication channels built on the Noise protocol framework. Pontis reduces attestation complexity from <em>O</em>(<em>n</em>²) → <em>O</em>(<em>n</em>), achieving up to 99% reduction in required attestations for large-scale deployments. Comprehensive evaluations on heterogeneous TEE platforms, including Intel SGX and ARM TrustZone, demonstrate that Pontis maintains an attestation latency of 60 ms with over 10,000 TEE instances, while establishing a trusted channel between heterogeneous TEEs requires only 5 ms. These results demonstrate the robustness and feasibility of Pontis, establishing a robust foundation for secure, scalable, and flexible cross-platform collaboration in demanding heterogeneous computing environments.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104652"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190593","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}

{"title":"Structure-aware multi-view urban representation learning with coordinated fusion and alignment","authors":"Jinghui Wei ,&nbsp;Sheng Wu ,&nbsp;Shifen Cheng ,&nbsp;Peixiao Wang ,&nbsp;Feng Lu","doi":"10.1016/j.ipm.2026.104672","DOIUrl":"10.1016/j.ipm.2026.104672","url":null,"abstract":"<div><div>Urban representation learning leverages heterogeneous data. While unified frameworks that combine feature fusion and contrastive learning have achieved promising results, two key challenges remain: 1) the lack of structural awareness often leads to suboptimal negative sampling and reduces the discriminability of embeddings; and 2) the inherently conflicting optimization objectives between fusion and contrastive modules may result in unstable training and suboptimal convergence. To address these issues, we propose SAMC, a <strong>S</strong>tructure-<strong>A</strong>ware <strong>M</strong>ulti-view representation learning framework with <strong>C</strong>oordinated fusion and alignment. SAMC introduces a multi-view contrastive learning module that incorporates structural similarity into negative sampling, thereby enhancing semantic coherence and cross-view consistency. To improve training stability, a flexible optimization strategy that incorporates soft Lagrangian constraints and stepwise state tracking is designed to coordinate gradient updates across fusion and alignment modules. Experiments on multiple urban datasets show that SAMC achieves average improvements of approximately 17.2%, 14.2%, and 5.9% compared to the state-of-the-art baselines on the tasks of regional popularity prediction, service demand prediction, and land use classification, respectively. Visualization analyses further confirm that SAMC enhances the discriminability, robustness, and generalizability of urban representations by embedding structural priors and adopting multi-stage coordinated optimization. Moreover, SAMC achieves a favorable trade-off between computational efficiency and predictive performance.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104672"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190571","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}

{"title":"Source-free time series domain adaptation with class-aware temporal imputation","authors":"Yingyi Zhong,&nbsp;Wen’an Zhou,&nbsp;Liwen Tao","doi":"10.1016/j.ipm.2026.104658","DOIUrl":"10.1016/j.ipm.2026.104658","url":null,"abstract":"<div><div>Source-free domain adaptation (SFDA) aims to adapt a pre-trained source model to the target domain without access to the source domain data. Recent time series SFDA works transfer temporal dynamics by employing a self-supervised temporal imputation approach designed to reconstruct masked sequences. However, this approach suffers from semantic ambiguity in the reconstructed input and task/classification-irrelevance in the reconstruction objective, which leads to the neglect of class-specific temporal details. As a result, the source model struggles to learn sufficiently discriminative temporal dynamics, limiting its performance on the target domain classification task. To address these issues, we propose Class-Aware Temporal Imputation (CATI), a novel time series SFDA method. CATI improves the existing temporal imputation approach during source pre-training by introducing a class-aware mechanism comprising two modules: Task-guided Contrastive Learning (TCL) and Gradient-based Semantic Enhancement (GSE). TCL refines the global semantic structure through contrastive learning to improve the semantic clarity of the reconstructed input, while GSE highlights local task-relevant regions through backpropagated gradients to guide the reconstruction process towards discriminative temporal details. By combining TCL and GSE, CATI enables the source model to better capture class-specific temporal dynamics, leading to improved adaptation performance on the target domain. Extensive experiments on four real-world time series datasets demonstrate that CATI surpasses state-of-the-art domain adaptation methods, with average gains of 1.41% in accuracy and 1.54% in MF1-score.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 5","pages":"Article 104658"},"PeriodicalIF":6.9,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190591","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}