Information Processing & Management最新文献

{"title":"Confusing negative commonsense knowledge generation with hierarchy modeling and LLM-enhanced filtering","authors":"Yaqing Sheng,&nbsp;Weixin Zeng,&nbsp;Jiuyang Tang,&nbsp;Lihua Liu,&nbsp;Xiang Zhao","doi":"10.1016/j.ipm.2025.104060","DOIUrl":"10.1016/j.ipm.2025.104060","url":null,"abstract":"<div><div>While most of the world’s knowledge exists in a positive and affirmative form, negative knowledge also plays a significant role by showing what is not true or what not to think, and has yet been largely overlooked. Existing negative commonsense knowledge generation methods adopt the generation-filtering paradigm, while the produced negative statements are easy to detect and fail to contribute to both human perception and task-specific algorithms that require negative samples for training. In response, we put forward <span>CONEG</span>, a negative commonsense knowledge generation framework that generates <em>confusing</em> statements, featuring hierarchy modeling in candidate generation and LLM-enhanced two-stage filtering. Specifically, in the <em>candidate generation</em> stage, we identify <em>congeners</em> for entity phrases in the commonsense knowledge base using box embeddings, which can effectively capture the hierarchical correlations among entity phrases and produce confusing candidates. In the <em>candidate filtering</em> stage, we design a two-stage filtering strategy, consisting of <em>intrinsic</em> triple confidence measuring and <em>extrinsic</em> refinement through large language models with group-based instructions, which can effectively filter out true facts and low-quality negative candidates. We empirically evaluate our proposal on both intrinsic assessment and downstream tasks, and the results demonstrate that <span>CONEG</span> and its components are effective in terms of producing confusing negative knowledge, surpassing the state-of-the-art methods.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104060"},"PeriodicalIF":7.4,"publicationDate":"2025-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139099","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":"Disentangled feature graph for Hierarchical Text Classification","authors":"Renyuan Liu,&nbsp;Xuejie Zhang,&nbsp;Jin Wang,&nbsp;Xiaobing Zhou","doi":"10.1016/j.ipm.2025.104065","DOIUrl":"10.1016/j.ipm.2025.104065","url":null,"abstract":"<div><div>Effectively utilizing the hierarchical relationship among labels is the core of Hierarchical Text Classification (HTC). Previous research on HTC has tended to enhance the dependencies between labels. However, they overlook some labels that may conflict with other labels because alleviating label conflicts also weakens label dependencies and reduces the model performance. Therefore, this paper focuses on the issue of label conflicts and studies methods to alleviate label conflicts without affecting the mutual support relationship between labels. To solve the abovementioned problem, we first use the feature disentanglement method to cut off all label connections. Then, the connection among labels is selectively established by constructing a hierarchical graph on disentangled features. Finally, the Graph Neural Networks (GNN) is adopted to encode the obtained Disentanglement Feature Graph (DFG) and enables only labels with connections to support each other, while labels without connections do not interfere with each other. The experimental results on the WOS, RCV1-v2, and BGC datasets show the effectiveness of DFG. In detail, the experimental results show that on the WOS dataset, the model incorporating DFG achieved a 1.07% improvement in Macro-F1, surpassing the best model by 0.27%. On the RCV1-v2 dataset, the model incorporating DFG achieved a 0.95% improvement in Micro-F1, surpassing the best model by 0.21%. On the BGC dataset, the model incorporating DFG achieved a 1.81% improvement in Micro-F1, surpassing the best model by 0.45%.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104065"},"PeriodicalIF":7.4,"publicationDate":"2025-01-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139098","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":"Why logit distillation works: A novel knowledge distillation technique by deriving target augmentation and logits distortion","authors":"Md Imtiaz Hossain,&nbsp;Sharmen Akhter,&nbsp;Nosin Ibna Mahbub,&nbsp;Choong Seon Hong,&nbsp;Eui-Nam Huh","doi":"10.1016/j.ipm.2024.104056","DOIUrl":"10.1016/j.ipm.2024.104056","url":null,"abstract":"<div><div>Although logit distillation aims to transfer knowledge from a large teacher network to a student, the underlying mechanisms and reasons for its effectiveness are unclear. This article explains the effectiveness of knowledge distillation (KD). Based on the observations, this paper proposes a novel distillation technique called TALD-KD that performs through Target Augmentation and a novel concept of dynamic Logits Distortion technique. The proposed TALD-KD unraveled the intricate relationships of dark knowledge semantics, randomness, flexibility, and augmentation with logits-level KD via three different investigations, hypotheses, and observations. TALD-KD improved student generalization through the linear combination of the teacher logits and random noise. Among the three versions assessed (TALD-A, TALD-B, and TALD-C), TALD-B improved the performance of KD on a large-scale ImageNet-1K dataset from 68.87% to 69.58% for top-1 accuracy, and from 88.76% to 90.13% for top-5 accuracy. Similarly, for the state-of-the-art approach, DKD, the performance improvements by the TALD-B ranged from 72.05% to 72.81% for top-1 accuracy and from 91.05% to 92.04% for top-5 accuracy. The other versions revealed the secrets of logit-level KD. Extensive ablation studies confirmed the superiority of the proposed approach over existing state-of-the-art approaches in diverse scenarios.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104056"},"PeriodicalIF":7.4,"publicationDate":"2025-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139102","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":"Psycholinguistic knowledge-guided graph network for personality detection of silent users","authors":"Houjie Qiu ,&nbsp;Xingkong Ma ,&nbsp;Bo Liu ,&nbsp;Yiqing Cai ,&nbsp;Xinyi Chen ,&nbsp;Zhaoyun Ding","doi":"10.1016/j.ipm.2025.104064","DOIUrl":"10.1016/j.ipm.2025.104064","url":null,"abstract":"<div><div>Personality detection is an emerging task benefiting numerous fields. The existing studies on text-based personality detection rarely concern silent users who never publish social texts due to the lack of their posts. Simultaneously, the mainstream methods lack an effective pathway for silent user representation to detect the personality. To solve the silent user problems, we propose a psycholinguistic knowledge-guided graph network, <em>PKGN</em>. Our method is composed of neighbor post metric, graph initialization &amp; learning, and classification. Under the guidance of psychological knowledge, our model first selects high-quality posts from neighbors as the posts of silent users through the neighbor post metric. In the graph initialization &amp; learning, psychologically relevant categories are introduced to build the bipartite graph for each silent user and obtain the user representation via GATv2. Then, we utilize linear classifiers for personality classification. We conducted extensive experiments on a new real-world dataset, including 1581 samples. To conduct a baseline benchmark for the silent user personality detection task, we apply the neighbor post metric to combine with the existing work. From the experimental results, our model achieves 64.11% average accuracy and 63.21% average macro-F1, outperforming mainstream methods in most individual personality traits and comprehensive comparisons. Furthermore, the introduction of psycholinguistic knowledge benefits the model performance. In neighbor post metric comparison, the psycholinguistic knowledge from LIWC reduces the standard variances of psychological category count and improves the detection results (3.01% <span><math><mi>↑</mi></math></span> for average accuracy and 2.34% <span><math><mi>↑</mi></math></span> for average macro-F1). In the ablation study, all the psychologically relevant categories contribute to the model performance (ranging from 0.02% <span><math><mi>↑</mi></math></span> to 3.02% <span><math><mi>↑</mi></math></span> for average macro-F1).</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104064"},"PeriodicalIF":7.4,"publicationDate":"2025-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139104","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 diversity-enhanced knowledge distillation model for practical math word problem solving","authors":"Yi Zhang ,&nbsp;Guangyou Zhou ,&nbsp;Zhiwen Xie ,&nbsp;Jinjin Ma ,&nbsp;Jimmy Xiangji Huang","doi":"10.1016/j.ipm.2025.104059","DOIUrl":"10.1016/j.ipm.2025.104059","url":null,"abstract":"<div><div>Math Word Problem (MWP) solving is a critical task in natural language processing, has garnered significant research interest in recent years. Various recent studies heavily rely on Seq2Seq models and their extensions (e.g., Seq2Tree and Graph2Tree) to generate mathematical equations. While effective, these models struggle to generate diverse but counterpart solution equations, limiting their generalization across various math problem scenarios. In this paper, we introduce a novel Diversity-enhanced Knowledge Distillation (DivKD) model for practical MWP solving. Our approach proposes an adaptive diversity distillation method, in which a student model learns diverse equations by selectively transferring high-quality knowledge from a teacher model. Additionally, we design a diversity prior-enhanced student model to better capture the diversity distribution of equations by incorporating a conditional variational auto-encoder. Extensive experiments on four MWP benchmark datasets demonstrate that our approach achieves higher answer accuracy than strong baselines while maintaining high efficiency for practical applications.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104059"},"PeriodicalIF":7.4,"publicationDate":"2025-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139105","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":"Adaptive Asymmetric Supervised Cross-Modal Hashing with consensus matrix","authors":"Yinan Li ,&nbsp;Jun Long ,&nbsp;Youyuan Huang ,&nbsp;Zhan Yang","doi":"10.1016/j.ipm.2024.104037","DOIUrl":"10.1016/j.ipm.2024.104037","url":null,"abstract":"<div><div>Supervised hashing has garnered considerable attention in cross-modal retrieval by programming annotated diverse modality data into the unified binary representation that facilitates efficient retrieval and lightweight storage. Despite its advantages, a major challenge remains, how to get the utmost out of annotated information and derive robust common representation that accurately preserves the intrinsic relations across heterogeneous modalities. In this paper, we present an innovative <strong>A</strong>daptive <strong>A</strong>symmetric <strong>S</strong>upervised <strong>C</strong>ross-modal <strong>H</strong>ashing method with consensus matrix to tackle the problem. We begin by formulating the proposition through matrix factorization to obtain the common representation utilizing consensus matrix efficiently. To safeguard the completeness of diverse modality data, we incorporate them via adaptive weight factors along with nuclear norms. Furthermore, an asymmetric hash learning framework between the representative coefficient matrices that come from common representation and semantic labels was constructed to constitute concentrated hash codes. Additionally, a valid discrete optimization algorithm was programmed. Comprehensive experiments conducted on MIRFlirck, NUS-WIDE, and IARP-TC12 datasets validate that <strong>A2SCH</strong> outperforms leading-edge hashing methods in cross-modal retrieval tasks.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104037"},"PeriodicalIF":7.4,"publicationDate":"2025-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139366","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":"Contrastive deep graph clustering with hard boundary sample awareness","authors":"Linlin Zhu,&nbsp;Heli Sun,&nbsp;Xiaoyong Huang,&nbsp;Pan Lou,&nbsp;Liang He","doi":"10.1016/j.ipm.2024.104050","DOIUrl":"10.1016/j.ipm.2024.104050","url":null,"abstract":"<div><div>Contrastive deep graph clustering is a graph data clustering method that combines deep learning and contrastive learning, aiming to realize accurate clustering of graph nodes. Although existing hard sample mining-based methods show positive results, they face the following challenges: (1) Clustering methods based on data similarity may lose important intrinsic groupings and association patterns, thereby affecting a comprehensive understanding and interpretation of the data. (2) In the measurement of hard samples, ignoring hard boundary samples may exacerbate clustering bias. To address these issues, we propose a new contrastive deep graph clustering method called Hard Boundary Sample Aware Network (HBSAN), which introduces attribute and structure enhanced encoding and generalized dynamic hard boundary sample weighting modulation strategy. Specifically, we optimize the similarity computation among samples through adaptive attribute embedding and multiview structure embedding techniques to deeply explore the intrinsic connections among samples, thereby aiding in the measurement of hard boundary samples. Furthermore, we leverage the unreliable confidence information obtained from initial clustering analysis to design an innovative hard boundary sample weight modulation function. This function first identifies the hard boundary samples and then dynamically reduces their weights, effectively enhancing the discriminative capability of network in ambiguous classification scenarios. Combining extensive experimental evaluations and in-depth analysis, our approach achieves state-of-the-art performance and establishes superior results in handling complex network clustering tasks.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104050"},"PeriodicalIF":7.4,"publicationDate":"2025-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139101","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":"Drivers of consumer group participation in an online shopping event: Alibaba's singles’ day","authors":"Jen-Her Wu ,&nbsp;Qi Li ,&nbsp;Lifang Peng ,&nbsp;Simon Robinson ,&nbsp;Yi-Cheng Chen","doi":"10.1016/j.ipm.2025.104067","DOIUrl":"10.1016/j.ipm.2025.104067","url":null,"abstract":"<div><div>Singles’ Day, an online shopping event (OSE), is a dynamic digital experience with consumer incentives, hedonic presence, and social gamification designed to encourage active social interaction. Based upon Activity theory, we identified intrinsic motivation, platform synergy, hedonic presence, and social gamification as key factors affecting consumer group participation during an OSE. Using 614 valid responses collected from an OSE, we reveal: (1) intrinsic motivation, platform synergy, and hedonic presence had direct effects on social gamification, which in turn affected consumer group participation, and (2) social gamification exerted significant mediating impacts on the influencing relationships between intrinsic motivation, platform synergy, hedonic presence, and group participation. These findings provide a nuanced understanding of the interplay between tools and group behavioural outcomes in OSEs. This study contributes a crucial extension of the in-depth understanding of the mechanisms by which elements of Activity theory and IT-enabled platforms individually and jointly enhance consumer group participation.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104067"},"PeriodicalIF":7.4,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139394","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 contracted container-based code component collaboration model with reusable but invisible right management","authors":"Wei Wang ,&nbsp;Zhenping Xie","doi":"10.1016/j.ipm.2024.104057","DOIUrl":"10.1016/j.ipm.2024.104057","url":null,"abstract":"<div><div>Existing methodologies for the code collaboration are facing challenges such as leakage of privacy data and insecure centralized management. To alleviate these challenges, we propose a contracted container-based code component collaboration model. Firstly, a specific blockchain is developed as a computational network that supports the storage and verification of sensitive data, which integrates an off-chain file system and encryption algorithms to realize the traceability and version control of privacy data. Secondly, the containerization and smart contract technology are introduced to enhance access control and the secure management and collaboration of code components. Thirdly, a protocol is designed to ensure the high reliability of cross-node information communication. Finally, the results of the prototype experiment demonstrate that the model effectively resists common attacks, meets critical security criteria, and maintains stable performance in the collaboration process. Moreover, compared to the state-of-the-art research, our model implements more valuable functionality characteristics, design goals and security attributes in the code component collaboration practice.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104057"},"PeriodicalIF":7.4,"publicationDate":"2025-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139393","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":"INSNER: A generative instruction-based prompting method for boosting performance in few-shot NER","authors":"Peiwen Zhao ,&nbsp;Chong Feng ,&nbsp;Peiguang Li ,&nbsp;Guanting Dong ,&nbsp;Sirui Wang","doi":"10.1016/j.ipm.2024.104040","DOIUrl":"10.1016/j.ipm.2024.104040","url":null,"abstract":"<div><div>Most existing Named Entity Recognition (NER) methods require a large scale of labeled data and exhibit poor performance in low-resource scenarios. Thus in this paper, we propose INSNER, a generative INStruction-based prompting method for few-shot NER. Specifically, we introduce a unified instruction to guide the model in extracting correct entities in response to the instruction, and construct synthetic verbalizers, which support complex types, to encourage effective knowledge transfer. We organize the NER results in natural language form, which mitigates the gap between pre-training and fine-tuning of language models. Furthermore, to facilitate the model to learn task-related knowledge and rich label semantics, we introduce entity-oriented prompt-tuning as an auxiliary task. We conduct in-domain and cross-domain experiments in few-shot settings on 4 datasets, and extensive analyses to validate the effectiveness and generalization ability of INSNER. Experimental results demonstrate that INSNER significantly outperforms current methods in few-shot settings, especially huge improvements(+12.0% F1) over the powerful ChatGPT in MIT Movie Complex both under a 10-shot setting.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"62 3","pages":"Article 104040"},"PeriodicalIF":7.4,"publicationDate":"2025-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143139391","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}