Information Processing & Management最新文献

{"title":"Modeling semantic representation with LLM-enhanced for knowledge-aware recommendation","authors":"Jianfang Liu ,&nbsp;Baolin Yi ,&nbsp;Huanyu Zhang ,&nbsp;Xiaoxuan Shen ,&nbsp;Lingling Song ,&nbsp;Yu Lei ,&nbsp;Hao Zheng","doi":"10.1016/j.ipm.2025.104387","DOIUrl":"10.1016/j.ipm.2025.104387","url":null,"abstract":"<div><div>Knowledge graph-based recommendation systems utilize structured entity and relation representations to better model user preferences. However, many traditional approaches rely primarily on ID-based data and often overlook textual information associated with items and relations, leading to limited semantic understanding. While recent approaches have begun incorporating large language models (LLMs), most focus solely on enhancing relational embeddings and fail to fully exploit the semantic extraction capabilities of LLMs. To address these limitations, we propose <em><u>LLMKnowRec</u></em>, a novel LLM-enhanced, knowledge-aware recommendation framework designed to improve the semantic modeling of knowledge graphs. Our approach integrates the powerful language understanding abilities of LLMs with traditional ID-based recommendation by introducing an LLM-based embedding generator. This generator produces semantically rich embeddings using textual descriptions of user profiles and knowledge graph relations. Building on this, we further introduce a semantic user intent modeling module, which leverages LLMs to aggregate multiple intent signals into comprehensive, semantically enriched intent embeddings. Additionally, we develop a relational intent-aware aggregation scheme that effectively combines higher-order representations, capturing both relational structures and user intent, thus enhancing the overall semantic understanding of users and items. Experimental conducted on three public benchmark datasets demonstrate that <em>LLMKnowRec</em> consistently outperforms state-of-the-art methods. Specifically, our model achieves improvements of up to 12.92%, 19.27%, and 8.23% in NDCG@10, and up to 13.41%, 15.62%, and 23.55% in Recall@10 across the three datasets, respectively. These results demonstrate the effectiveness and practical potential of our proposed method. The implementation code is publicly available at: <span><span>https://github.com/liujianfang2021/LLMKnowRec</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104387"},"PeriodicalIF":6.9,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145027197","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":"Whole dataset context-aware prediction on the null values in time series data for faster inferencing with low complexity","authors":"Sharmen Akhter ,&nbsp;Nosin Ibna Mahbub ,&nbsp;Junyoung Park ,&nbsp;Eui-Nam Huh","doi":"10.1016/j.ipm.2025.104370","DOIUrl":"10.1016/j.ipm.2025.104370","url":null,"abstract":"<div><div>Null value handling in time series datasets demands heavy and explicit data imputation models, which add latency during inferences for various tasks. This concern raises the following question: <em>is it possible to avoid these explicit data imputation models to perform task predictions directly during inference without imputation?</em> As a pioneer, this paper proposes <strong>ZeroTIP</strong>, a knowledge distillation (KD)-based <strong>Zero T</strong>ime <strong>I</strong>mputation for <strong>P</strong>rediction strategy for workload predictions without having additional data-imputation models. During the training period, a student network is forced to reason the missing or null values implicitly and mimic the inference (workload prediction task) while taking synthetically corrupted data as input and being supervised by the pretrained teacher network that contains representations of the original dataset. Only the student network is used during inference. ZeroTIP reduced the inference time by almost 99.9% by avoiding explicit data imputation. A version of ZeroTIP, called ZeroTIP-DI, was deployed for the data imputation task to evaluate the significance of ZeroTIP in reasoning data context and pattern. For a prediction length of 48 and 96, ZeroTIP-DI achieved an average improvement of 38.37 (97.08%) and 21.67 (95.08%) times the baseline, highlighting its superiority.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104370"},"PeriodicalIF":6.9,"publicationDate":"2025-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145020340","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":"Preserving overlapped information via parallel one-hop and multi-hop neighbor encoding for knowledge graph entity typing","authors":"Hongbin Zhang ,&nbsp;Zhenghao Huang ,&nbsp;Ruihao Li ,&nbsp;Tao Wang ,&nbsp;Zhuowei Wang ,&nbsp;Lianglun Cheng","doi":"10.1016/j.ipm.2025.104385","DOIUrl":"10.1016/j.ipm.2025.104385","url":null,"abstract":"<div><div>Knowledge graph entity typing (KGET) is critical for predicting and completing missing entity types in a knowledge graph (KG). Existing KGET models mainly aggregate local semantic and structural information from multi-hop and one-hop neighbors via weighted aggregation. However, the one-hop neighbor information within the multi-hop neighbor context is often diluted during aggregation, resulting in incomplete information collection and inaccurate type prediction. To preserve this overlapped one-hop neighbor information, we propose a novel framework, the one-hop and multi-hop neighbor parallel encoding framework (OMNPEF), which captures local-to-global semantic and structural information. Specifically, OMNPEF encodes one-hop and multi-hop neighbors in parallel to better preserve the overlapped one-hop neighbor information and integrates local information with global semantic and structural insights, enhancing the model’s capacity to learn from the graph structure. Experimental results on the FB15kET and YAGO43kET datasets demonstrate that OMNPEF outperforms state-of-the-art models, achieving a mean improvement of at least 1.3% in MRR.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104385"},"PeriodicalIF":6.9,"publicationDate":"2025-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145020343","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":"Distilling knowledge from large language models: A concept bottleneck model for hate and counter speech recognition","authors":"Roberto Labadie-Tamayo ,&nbsp;Djordje Slijepčević ,&nbsp;Xihui Chen ,&nbsp;Adrian Jaques Böck ,&nbsp;Andreas Babic ,&nbsp;Liz Freimann ,&nbsp;Christiane Atzmüller ,&nbsp;Matthias Zeppelzauer","doi":"10.1016/j.ipm.2025.104309","DOIUrl":"10.1016/j.ipm.2025.104309","url":null,"abstract":"<div><div>The rapid increase in hate speech on social media has exposed an unprecedented impact on society, making automated methods for detecting such content important. Unlike prior black-box models, we propose a novel transparent method for automated hate and counter speech recognition, i.e., “Speech Concept Bottleneck Model” (SCBM), using adjectives as human-interpretable bottleneck concepts. SCBM leverages large language models (LLMs) to map input texts to an abstract adjective-based representation, which is then sent to a light-weight classifier for downstream tasks. Across five benchmark datasets spanning multiple languages and platforms (e.g., Twitter, Reddit, YouTube), SCBM achieves an average macro-F1 score of 0.69 which outperforms the most recently reported results from the literature on four out of five datasets. Aside from high recognition accuracy, SCBM provides a high level of both local and global interpretability. Furthermore, fusing our adjective-based concept representation with transformer embeddings, leads to a 1.8% performance increase on average across all datasets, showing that the proposed representation captures complementary information. Our results demonstrate that adjective-based concept representations can serve as compact, interpretable, and effective encodings for hate and counter speech recognition. With adapted adjectives, our method can also be applied to other NLP tasks.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104309"},"PeriodicalIF":6.9,"publicationDate":"2025-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145020341","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":"Emotion-semantic interaction network for fake news detection: Perspectives on question and non-question comment semantics","authors":"Zhenhua Tan ,&nbsp;Tao Zhang","doi":"10.1016/j.ipm.2025.104391","DOIUrl":"10.1016/j.ipm.2025.104391","url":null,"abstract":"<div><div>Current fake news detectors often overlook the association between news content emotion and comment semantics, especially questioning vs. non-questioning language. We observe that fake news evokes distinct comment patterns: non-sad emotions (e.g., anger, surprise) in fake content drive more questioning semantics (e.g., “Where is the video to support this?”) and suppress non-questioning replies, while sadness shows the opposite trend. Ignoring this emotion-semantic interaction limits detection accuracy. To address this limitation, we propose an Emotion-Semantic Interaction Network (ESIN), which learns a latent association between content emotions and comment semantics from questioning and non-questioning perspectives. Specifically, ESIN incorporates the interaction of content emotion with original comment semantics, as well as the distribution of extracted comment semantic categories (i.e., question and non-question), achieved through our proposed mechanisms called Codebook Initialization and Semantic Quantification, based on cross-attention. The ESIN model is comprehensively evaluated on two widely used fake news datasets, namely RumourEval and WeiBo. The ESIN achieves competitive performance, outperforming baseline models by a significant margin of 5.27% and 4.08% in weighted F1-score (F1.) and accuracy (Acc.) on the RumourEval dataset, and by 2.43% and 2.38% in F1. and Acc. on the WeiBo dataset. The promising result verifies the effectiveness of our ESIN model. Theoretically, this study fills a critical gap by highlighting how content emotions shape comment semantics (questioning vs. non-questioning) as a veracity cue, advancing understanding of user engagement patterns in fake news spread. Practically, ESIN offers actionable strategies for platforms to flag high-risk content and equips policymakers with tools for user training and evidence-based regulations, enhancing misinformation mitigation and public trust in digital information ecosystems.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104391"},"PeriodicalIF":6.9,"publicationDate":"2025-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145020342","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":"Heterogeneous data-driven resolution generation for software systems via large language models","authors":"Wen Liu ,&nbsp;Degang Sun ,&nbsp;Haitian Yang ,&nbsp;Yan Wang ,&nbsp;Weiqing Huang","doi":"10.1016/j.ipm.2025.104376","DOIUrl":"10.1016/j.ipm.2025.104376","url":null,"abstract":"<div><div>Modern software systems are increasingly complex and dynamic, making them particularly vulnerable to performance anomalies. Although runtime anomaly detection enhances system reliability, engineers still devote considerable time and effort to resolving errors once anomalous logs or metrics are detected. Such challenges call for intelligent automation capable of delivering targeted remediation steps based on detected anomalies. In this work, we first construct an anomaly-related knowledge base by combining heterogeneous operational data, including logs and metrics, with resolutions annotated by domain experts. Furthermore, we propose HASolver, the first Heterogeneous Anomaly Solver to generate recommended resolutions for multi-source system anomalies. The core component is a dual-view multi-vector module, designed to represent heterogeneous anomaly chunks from different modalities and to support effective multi-vector retrieval. HASolver integrates a large language model with domain knowledge to generate mitigation resolutions. We conduct extensive experiments using BLEU and ROUGE-1/2/L metrics. Compared to baseline approaches, HASolver delivers notable performance gains, improving BLEU and ROUGE-L scores by 14.6% and 19.6%, respectively. Further analyses are carried out to explore various multi-vector configurations and the effect of prompt strategies. We also release the annotated resolution dataset derived from the anomaly-related knowledge base to facilitate future research.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104376"},"PeriodicalIF":6.9,"publicationDate":"2025-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145020344","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 cost-effective community-hierarchy-based mutual voting approach for influence maximization in complex networks","authors":"Yi Liu ,&nbsp;Xiaoan Tang ,&nbsp;Witold Pedrycz ,&nbsp;Qiang Zhang","doi":"10.1016/j.ipm.2025.104371","DOIUrl":"10.1016/j.ipm.2025.104371","url":null,"abstract":"<div><div>Influence maximization seeks to choose influential nodes that can spread influence most widely in complex networks. However, current methods often fail to balance the accuracy of selecting such nodes with computational efficiency. To address this challenge, this article proposes a novel approach called Cost-Effective Community-Hierarchy-Based Mutual Voting for influence maximization in complex networks. First, we develop a method for measuring the importance of different nodes in networks based on an original concept of Dual-Scale Community-Hierarchy Information that synthesizes both hierarchy structural information and community structural information of nodes. The community structural information contained in the nodes is measured by a new notion of Hierarchical-Community Entropy. Second, we develop a method named Cost-Effective Mutual-Influence-based Voting for seed nodes selection. Hereinto, a low-computational-cost mutual voting mechanism and an updating strategy called Lazy Score Updating Strategy are newly constructed for optimizing the selecting of seed nodes. Third, we develop a balance index to evaluate the performance of different methods in striking the tradeoff between time complexity and the accuracy of influential nodes identification. Based on this index, we further propose a balance gap to quantify the distance between each method and the best achievable trade-off. Finally, we demonstrate the effectiveness of the proposed approach in terms of time complexity and spreading capability by comparing the experimental results based on five criteria over 13 public datasets. The extensive experiments show that the proposed approach outperforms 16 state-of-the-art techniques on the balance between time complexity and accuracy of influential nodes identification. Compared with the method that has the second highest mean Balance Index, our approach shows an improvement of up to 9.87 %, with the lowest improvement being 5.09 %, and an average improvement of 7.30 %. Moreover, our method consistently reaches the optimal balance point, as indicated by a mean Balance Gap value of zero across all networks and scenarios.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104371"},"PeriodicalIF":6.9,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145010849","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":"Meeting companies’ innovative requirements on online technology trading platforms: A novel large language model-based framework","authors":"Qingyu Xu,&nbsp;Zhaobin Liu,&nbsp;Jian Ma","doi":"10.1016/j.ipm.2025.104392","DOIUrl":"10.1016/j.ipm.2025.104392","url":null,"abstract":"<div><div>Online technology trading platforms (OTTPs) are critical for companies to publish technology requirements and identify solutions like patents. However, semantic gaps persist between market-driven needs and technical supply texts, which traditional methods fail to bridge. While large language models (LLMs) show promise, their effectiveness in OTTPs is limited by hallucination and temporal unawareness. We propose an LLM framework integrating the Hypothetical Document Embedding (HyDE) framework, where we generate pseudo-supply texts based on technical requirements. These texts are then matched with candidate patents using similarity calculations. To reduce hallucination, we use industry-specific knowledge graphs to guide the text generation process and introduce a self-reflective mechanism to refine the generated texts. To address the lack of time awareness, we enhance the knowledge graph with timestamps, turning it into a temporal knowledge graph. Additionally, we introduce the TPPR (Temporal Personalized PageRank) algorithm to improve the relevance of generated texts. Experiments show that our framework performs better than existing methods in Recall, Precision, and Mean Reciprocal Rank (MRR). This framework advances technology forecasting by enabling dynamic patent matching, offering organizations actionable insights for R&amp;D investments. ​By reducing mismatches and innovation cycle times, it supports sustainable technology transfer—highlighting implications for AI governance in evolving innovation ecosystems.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104392"},"PeriodicalIF":6.9,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145010927","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":"Quality of human-GenAI collaboration and its driving factors: A symbiotic agency perspective","authors":"Jiayu Shang ,&nbsp;Dan Huang ,&nbsp;Songshan (Sam) Huang","doi":"10.1016/j.ipm.2025.104373","DOIUrl":"10.1016/j.ipm.2025.104373","url":null,"abstract":"<div><div>Generative AI (GenAI) is increasingly integrated into users’ daily work as a collaborative partner. Drawing on the symbiotic agency theory, this study investigates the quality of human-GenAI collaboration using a mixed-methods approach, including qualitative interviews and quantitative surveys. Study 1 identified three dimensions of human-GenAI collaboration quality that comprise outcome quality, comfort, and efficiency; and six driving factors which can be categorized under human agency (domain knowledge, desire for control, and domestication ability), and GenAI agency (communication ability, working memory, and long-term memory). Study 2 applied fuzzy-set qualitative comparative analysis (fsQCA) to explore configurations of the driving factors that lead to high collaboration quality. Four distinct configurations emerged: 1) Domain knowledge, domestication ability, communication ability, and working memory; 2) Domain knowledge, domestication ability, working memory, and long-term memory; 3) Domain knowledge, communication ability, working memory, and long-term memory; and 4) Domain knowledge, ∼desire for control, domestication ability, communication ability, and long-term memory. These results advance understanding of human-GenAI collaboration by highlighting critical configurations of human and GenAI agency that foster high-quality collaboration. The study offers actionable insights to enhance human-GenAI interactions by optimizing both human and GenAI capabilities.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104373"},"PeriodicalIF":6.9,"publicationDate":"2025-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145004536","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":"MUSE-Rec: Explainable multi-behavioral social e-commerce recommendation with integrated link prediction","authors":"Gehad Abdullah Amran ,&nbsp;Xianneng Li ,&nbsp;Ali A. AL-Bakhrani","doi":"10.1016/j.ipm.2025.104355","DOIUrl":"10.1016/j.ipm.2025.104355","url":null,"abstract":"<div><div>Social e-commerce platforms create complex ecosystems where user behaviors, social relationships, and temporal dynamics are closely interconnected and evolve continuously over time. Current recommendation approaches face critical limitations: they struggle to model diverse user behaviors simultaneously, fail to predict evolving social connections, and lack interpretable explanations. Unlike existing methods that treat multi-behavioral modeling, social influence, and temporal dynamics as separate optimization problems, this work introduces MUSE-Rec, a unified Multi-behavioral Social E-commerce Recommendation Framework. MUSE-Rec integrates these interconnected components through an innovative graph neural network architecture. Integrating link prediction is crucial because predicting future social connections enables the system to anticipate how user preferences will evolve, improving recommendation accuracy and timing. Our framework advances recommendation systems theory by demonstrating that joint optimization of behavioral patterns, social dynamics, and temporal evolution achieves superior performance compared to component-wise approaches. This establishes new theoretical foundations for integrated social-temporal-behavioral modeling. MUSE-Rec introduces three key innovations: (1) a Multi-Graph Attention Network layer modeling diverse user-item interactions while predicting future social connections, achieving behavior correlation coefficient of 0.73 and link prediction AUC of 0.892; (2) an adaptive social connection aggregation mechanism capturing dynamic social influence patterns; and (3) a temporal graph network layer incorporating behavior-specific temporal dynamics. Comprehensive experiments on Yelp and Amazon Electronics datasets demonstrate superior performance. MUSE-Rec achieves NDCG@10 of 0.768 on Yelp and 0.742 on Amazon. The explainability module achieves high fidelity scores of 0.823 and 0.805 respectively, providing transparent behavior-specific explanations. MUSE-Rec enables e-commerce platforms to deploy more effective recommendation systems with 28% computational efficiency improvement while enhancing user trust.</div></div>","PeriodicalId":50365,"journal":{"name":"Information Processing & Management","volume":"63 2","pages":"Article 104355"},"PeriodicalIF":6.9,"publicationDate":"2025-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144997366","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}