Volume 2 · Issue 6 (2025)
Navigating Ambiguity, Strengthening Trust: Semantic Clarity and Relational Dynamics in Industry–University Knowledge Co-Creation
Qingyu Xu¹, Jian Ma1, *, Zhouqiang Qiu2
1 Department of Information System, City university of Hong Kong, China
2 Guangdong Science and Technology Innovation Monitoring and Research Center, Guangzhou, China
Corresponding Author: Jian Ma
Abstract: Effective university–industry collaboration remains challenging, primarily because companies, particularly small and medium-sized enterprises (SMEs), often struggle to clearly articulate their multi-dimensional technological requirements and accurately identify trustworthy academic experts aligned with these needs. To address these critical issues, we introduce DualRAG-SNR, a novel hierarchical matching framework explicitly designed to resolve semantic ambiguities and integrate trust signals. DualRAG-SNR incorporates (i) a dual-stage retrieval-augmented generation (DualRAG) mechanism, comprising a typed Graph-based retrieval (GraphRAG) followed by a precise Vector Database (VDB) retrieval, explicitly clarifying multi-aspect corporate requirements and retrieving semantically coherent knowledge; and (ii) an enriched Social Network Ranker (SNR) explicitly constructed from citation data, institutional affiliations, and fine-grained company–scholar interactions logged on the Online Technology Trading Platform (OTTP), capturing exploratory interactions such as profile views, communication exchanges, and contractual activities. RotatE embeddings explicitly model relational trust within this enriched social network. We further aggregate retrieved insights into a unified hypothetical requirement document (HyDE) using a large language model (LLM), explicitly enhancing semantic clarity. Through requirement-aware attention-based fusion, DualRAG-SNR dynamically balances semantic relevance and relational trust, significantly improving scholar recommendations. Empirical evaluation on real-world OTTP collaboration cases demonstrates that DualRAG-SNR achieves superior recall and nDCG compared to strong baselines. Furthermore, an interaction-based evaluation explicitly indicates that DualRAG-SNR recommendations elicit deeper and more sustained company–scholar interactions, explicitly signaling enhanced trust-building and more effective knowledge transfer. Ablation studies explicitly confirm that the hierarchical DualRAG retrieval, enriched exploratory interactions, and requirement-aware fusion contribute substantial complementary improvements. The proposed framework explicitly provides both theoretical insights and practical tools for systematically reducing information asymmetry and fostering robust knowledge transfer in university–industry partnerships.
Keywords: Retrieval-Augmented Generation, Social Network, University–Industry Collaboration, Knowledge Transfer, Text Matching
References
[1]
Albats E, Alexander AT, Cunningham JA (2022) Traditional, virtual, and digital intermediaries in university-industry collaboration: exploring institutional logics and bounded rationality. Technological Forecasting and Social Change 177:121470. doi: 10.1016/j.techfore.2022.121470
[2]
Ankrah S, AL-Tabbaa O (2015) Universities–industry collaboration: A systematic review. Scandinavian Journal of Management 31:387–408. doi: 10.1016/j.scaman.2015.02.003
[3]
Bikard M, Vakili K, Teodoridis F (2019) When Collaboration Bridges Institutions: The Impact of University–Industry Collaboration on Academic Productivity. Organization Science 30:426–445. doi: 10.1287/orsc.2018.1235
[4]
Bruneel J, D’Este P, Salter A (2010) Investigating the factors that diminish the barriers to university–industry collaboration. Research Policy 39:858–868. doi: 10.1016/j.respol.2010.03.006
[5]
Chen Y, Esmaeilzadeh P (2024) Generative AI in Medical Practice: In-Depth Exploration of Privacy and Security Challenges. J Med Internet Res 26:e53008. doi: 10.2196/53008
[6]
Cohen M, Fernandes G, Godinho P (2025) Measuring the impacts of university-industry R&D collaborations: a systematic literature review. J Technol Transf 50:345–374. doi: 10.1007/s10961-024-10114-5
[7]
He X, Meng X, Wu Y, Chan CS, Pang T (2020) Semantic Matching Efficiency of Supply and Demand Texts on Online Technology Trading Platforms: Taking the Electronic Information of Three Platforms as an Example. Information Processing & Management 57:102258. doi: 10.1016/j.ipm.2020.102258
[8]
Huang X, Lian J, Lei Y, Yao J, Lian D, Xie X (2025) Recommender AI Agent: Integrating Large Language Models for Interactive Recommendations. ACM Trans Inf Syst 43:1–33. doi: 10.1145/3731446
[9]
Ji Z, Lee N, Frieske R, Yu T, Su D, Xu Y, Ishii E, Bang YJ, Madotto A, Fung P (2023) Survey of Hallucination in Natural Language Generation. ACM Comput Surv 55:1–38. doi: 10.1145/3571730
[10]
Lewis P, Perez E, Piktus A, Petroni F, Karpukhin V, Goyal N, Küttler H, Lewis M, Yih W, Rocktäschel T, Riedel S, Kiela D (2020) Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. In: Larochelle H, Ranzato M, Hadsell R, Balcan MF, Lin H (eds) Advances in Neural Information Processing Systems. Curran Associates, Inc., pp 9459–9474
[11]
Li Y, Wang R, Nan G, Li D, Li M (2021) A personalized paper recommendation method considering diverse user preferences. Decision Support Systems 146:113546. doi: 10.1016/j.dss.2021.113546
[12]
Ligabue PDM, Brandão AAF, Peres SM, Cozman FG, Pirozelli P (2024) Applying a Context-based Method to Build a Knowledge Graph for the Blue Amazon. Data Intelligence 6:64–103. doi: 10.1162/dint_a_00223
[13]
Liu S, Cheng H, Wang Y, He Y, Fan C, Liu Z (2025) EvoPath: Evolutionary meta-path discovery with large language models for complex heterogeneous information networks. Information Processing & Management 62:103920. doi: 10.1016/j.ipm.2024.103920
[14]
Lyu S, Liu J (2021) Convolutional Recurrent Neural Networks for Text Classification. Journal of Database Management 32:65–82. doi: 10.4018/jdm.2021100105
[15]
Mi L, Huang L, Han Z, Miao H, Wu F (2022) Forecasting and evaluating emerging technologies based on supply and demand matching – a case study of China’s gerontechnology. Technology Analysis & Strategic Management 34:290–306. doi: 10.1080/09537325.2021.1895982
[16]
Pan S, Luo L, Wang Y, Chen C, Wang J, Wu X (2024) Unifying Large Language Models and Knowledge Graphs: A Roadmap. IEEE Trans Knowl Data Eng 36:3580–3599. doi: 10.1109/TKDE.2024.3352100
[17]
Plewa C, Korff N, Johnson C, Macpherson G, Baaken T, Rampersad GC (2013) The evolution of university–industry linkages—A framework. Journal of Engineering and Technology Management 30:21–44. doi: 10.1016/j.jengtecman.2012.11.005
[18]
Raiaan MAK, Mukta MdSH, Fatema K, Fahad NM, Sakib S, Mim MMJ, Ahmad J, Ali ME, Azam S (2024) A Review on Large Language Models: Architectures, Applications, Taxonomies, Open Issues and Challenges. IEEE Access 12:26839–26874. doi: 10.1109/access.2024.3365742
[19]
Reimers N, Gurevych I (2019) Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. In: Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). Association for Computational Linguistics, Hong Kong, China
[20]
Rybnicek R, Königsgruber R (2019) What makes industry–university collaboration succeed? A systematic review of the literature. J Bus Econ 89:221–250. doi: 10.1007/s11573-018-0916-6
[21]
Shanavas N, Wang H, Lin Z, Hawe G (2021) Knowledge-driven graph similarity for text classification. Int J Mach Learn & Cyber 12:1067–1081. doi: 10.1007/s13042-020-01221-4
[22]
Song M, Zhao W, HaiHong E (2020) KGAnet: a knowledge graph attention network for enhancing natural language inference. Neural Comput & Applic 32:14963–14973. doi: 10.1007/s00521-020-04851-5
[23]
Sun Z, Deng Z-H, Nie J-Y, Tang J RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space. In: International Conference on Learning Representations
[24]
Teng H, Wang N, Zhao H, Hu Y, Jin H (2024) Enhancing semantic text similarity with functional semantic knowledge (FOP) in patents. Journal of Informetrics 18:101467. doi: 10.1016/j.joi.2023.101467
[25]
Wang J, Pan M, He T, Huang X, Wang X, Tu X (2020) A Pseudo-relevance feedback framework combining relevance matching and semantic matching for information retrieval. Information Processing & Management 57:102342. doi: 10.1016/j.ipm.2020.102342
[26]
Wang Y, Wang L, Zhou Q, Wang Z, Li H, Hua G, Tang W (2024) Multimodal LLM Enhanced Cross-lingual Cross-modal Retrieval. In: Proceedings of the 32nd ACM International Conference on Multimedia. ACM, Melbourne VIC Australia, pp 8296–8305
[27]
Yu C, Xue H, Jiang Y, An L, Li G (2021) A simple and efficient text matching model based on deep interaction. Information Processing & Management 58:102738. doi: 10.1016/j.ipm.2021.102738
Download PDF