|
|
|
| Patent Transaction Recommendation via Integrating Large Language Models and Graph Representation Learning |
| Zhang Zhihao1, Huo Yilin2, Yuan Jiaxu1, He Xijun1 |
1.College of Economics and Management, Beijing University of Technology, Beijing 100124
2.School of Economics and Management, Beijing Jiaotong University, Beijing 100044 |
|
|
|
|
Abstract Patent transaction recommendation is a crucial pathway for promoting the transformation of scientific and technological achievements. However, it faces dual challenges: highly sparse transaction data and the difficulty in effectively modeling the complex interaction relationships between organizations and patents. To address these issues, this study proposes LLM-GRL-PTR, a novel patent transaction recommendation (PTR) model that integrates large language models (LLMs) and graph representation learning (GRL). First, to mitigate data sparsity, we proposed an LLM-based data augmentation method that leveraged LLMs to infer organizational demand preferences from sparse historical transaction records and screened potential matching patents, thereby enabling personalized data completion for long-tail organizations. Second, we constructed a GRL-based recommendation framework that jointly modeled three types of relationships: enhanced organization-patent interactions, multidimensional proximities among organizations, and upstream-downstream technical linkages among patents—reflecting the industrial chain synergy. Accordingly, we built three graphs depicting heterogeneous organization-patent interaction, interorganizational proximity, and patent upstream-downstream relationship. By integrating graph neural networks with attention mechanisms, the model fused explicit interaction signals and implicit relational features to learn low-dimensional representations of organizations and patents to ultimately predict potential transaction likelihoods. Finally, empirical studies were conducted in two critical domains—electronic information and chip manufacturing. The results showed that the proposed model significantly outperformed baseline methods in terms of recommendation accuracy, and the recommended patents exhibited broad technical coverage, effectively balancing accuracy and diversity. This study provides an effective tool for intelligent patent transaction recommendation and demonstrates a promising paradigm for integrating large models into technology intelligence services.
|
|
Received: 30 September 2025
|
|
|
|
1 曹春方, 龚曼宁. 标准定则市场兴——技术标准对专利交易的促进作用研究[J]. 管理世界, 2025, 41(1): 51-66, 107.
2 国家知识产权局. 2025年中国专利调查报告[R/OL]. (2026-04-01). https://www.cnipa.gov.cn/module/download/down.jsp?i_ID=205590colID=88.
3 国家知识产权局. 2024年中国专利调查报告[R/OL]. (2025-01-22). https://www.cnipa.gov.cn/art/2025/1/22/art_88_197320.html.
4 Liu Y, Li K W. A two-sided matching decision method for supply and demand of technological knowledge[J]. Journal of Knowledge Management, 2017, 21(3): 592-606.
5 何喜军, 石安杰, 吴爽爽, 等. 基于知识图谱与强化学习的专利交易推荐研究[J]. 系统工程理论与实践, 2024, 44(10): 3330-3345.
6 冉从敬, 宋凯. 基于混合方法的高校专利个性化推荐模型构建[J]. 情报理论与实践, 2020, 43(10): 93-98.
7 Zhang W X, Liu H Z, Dong Z J, et al. Bridging the information gap between domain-specific model and general LLM for personalized recommendation[C]// Proceedings of the 8th Aisa-Pacific Web and Web-Age Information Management Joint Conference on Web and Big Data. Singapore: Springer, 2024: 280-294.
8 Zhang D K, Yin J, Zhu X Q, et al. Network representation learning: a survey[J]. IEEE Transactions on Big Data, 2020, 6(1): 3-28.
9 Zhang Y, Qian Y, Huang Y, et al. An entropy-based indicator system for measuring the potential of patents in technological innovation: rejecting moderation[J]. Scientometrics, 2017, 111(3): 1925-1946.
10 姜南, 李济宇, 顾文君. 技术宽度、技术深度和知识转移[J]. 科学学研究, 2020, 38(9): 1638-1646.
11 何喜军, 石安杰, 武玉英, 等. 组态视角下基于TOE的可转化专利特征因果推断[J]. 科学学研究, 2022, 40(11): 1979-1990.
12 何喜军, 石安杰, 吴爽爽, 等. 基于Logit模型的技术供需交易影响因素分析[J]. 科技管理研究, 2024, 44(7): 80-86.
13 刘晓燕, 李金鹏, 单晓红, 等. 多维邻近性对集成电路产业专利技术交易的影响[J]. 科学学研究, 2020, 38(5): 834-842, 960.
14 李纲, 余辉, 梁镇涛, 等. 技术交易中供需匹配影响因素研究——基于TOE框架的组态分析[J]. 情报理论与实践, 2022, 45(2): 85-93, 120.
15 马荣康, 刘凤朝. 基于专利许可的新能源技术转移网络演变特征研究[J]. 科学学与科学技术管理, 2017, 38(6): 65-76.
16 刘凤朝, 肖站旗, 马荣康. 多维邻近性对技术交易网络的动态影响研究[J]. 科学学研究, 2018, 36(12): 2205-2214.
17 Ferretti M, Guerini M, Panetti E, et al. The partner next door? The effect of micro-geographical proximity on intra-cluster inter-organizational relationships[J]. Technovation, 2022, 111: 102390.
18 Resnick P, Varian H R. Recommender systems[J]. Communications of the ACM, 1997, 40(3): 56-58.
19 Trappey A J C, Trappey C V, Wu C Y, et al. Intelligent patent recommendation system for innovative design collaboration[J]. Journal of Network and Computer Applications, 2013, 36(6): 1441-1450.
20 Ji X, Gu X J, Dai F, et al. Patent collaborative filtering recommendation approach based on patent similarity[C]// Proceedings of the Eighth International Conference on Fuzzy Systems and Knowledge Discovery. Piscataway: IEEE, 2011: 1699-1703.
21 Jiang L, Yang C C. User recommendation in healthcare social media by assessing user similarity in heterogeneous network[J]. Artificial Intelligence in Medicine, 2017, 81: 63-77.
22 何喜军, 董艳波, 武玉英, 等. 基于异构信息网络嵌入的专利技术主体间交易推荐模型[J]. 情报学报, 2020, 39(1): 57-67.
23 Sun Y Z, Han J W. Integrating clustering with ranking in heterogeneous information networks analysis[M]// Link Mining: Models, Algorithms, and Applications. New York: Springer, 2010: 439-473.
24 Wang Q, Du W, Ma J, et al. Recommendation mechanism for patent trading empowered by heterogeneous information networks[J]. International Journal of Electronic Commerce, 2019, 23(2): 147-178.
25 He X J, Dong Y B, Zhen Z, et al. Weighted meta paths and networking embedding for patent technology trade recommendations among subjects[J]. Knowledge-Based Systems, 2019, 184: 104899.
26 Deng W W, Ma J. A knowledge graph approach for recommending patents to companies[J]. Electronic Commerce Research, 2022, 22(4): 1435-1466.
27 何喜军, 董颖, 张佑, 等. 基于供需知识图谱的燃料电池领域专利交易推荐[J]. 系统工程, 2025, 43(3): 13-25.
28 赵展一, 钟永恒, 王辉, 等. 基于技术关联关系的企业研发潜在合作伙伴识别方法研究综述[J]. 现代情报, 2023, 43(10): 152-163, 177.
29 Liu Y X, Zhang W N, Chen Y F, et al. Conversational recommender system and large language model are made for each other in E-commerce pre-sales dialogue[C]// Findings of the Association for Computational Linguistics: EMNLP 2023. Stroudsburg: Association for Computational Linguistics, 2023: 9587-9605.
30 Wu C H, Wu F Z, Qi T, et al. Empowering news recommendation with pre-trained language models[C]// Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM Press, 2021: 1652-1656.
31 Wu L K, Qiu Z P, Zheng Z, et al. Exploring large language model for graph data understanding in online job recommendations[C]// Proceedings of the AAAI Conference on Artificial Intelligence. Palo Alto: AAAI Press, 2024, 38(8): 9178-9186.
32 Hou Y P, Zhang J J, Lin Z H, et al. Large language models are zero-shot rankers for recommender systems[C]// Proceedings of the 46th European Conference on Information Retrieval. Cham: Springer, 2024: 364-381.
33 Bao K Q, Zhang J Z, Zhang Y, et al. TALLRec: an effective and efficient tuning framework to align large language model with recommendation[C]// Proceedings of the 17th ACM Conference on Recommender Systems. New York: ACM Press, 2023: 1007-1014.
34 张志豪, 刘思航, 马天骥, 等. 大模型驱动下基于企业画像的专利交易个性化推荐研究[J]. 情报理论与实践, 2025, 48(5): 177-186.
35 Harte J, Zorgdrager W, Louridas P, et al. Leveraging large language models for sequential recommendation[C]// Proceedings of the 17th ACM Conference on Recommender Systems. New York: ACM Press, 2023: 1096-1102.
36 Xi Y J, Liu W W, Lin J H, et al. Towards open-world recommendation with knowledge augmentation from large language models[C]// Proceedings of the 18th ACM Conference on Recommender Systems. New York: ACM Press, 2024: 12-22.
37 白如江, 陈启明, 张玉洁, 等. 基于ChatGPT+Prompt的专利技术功效实体自动生成研究[J]. 数据分析与知识发现, 2024, 8(4): 14-25.
38 刘奕涵, 石安杰, 李振威, 等. 粤港澳大湾区专利合作网络结构及链路预测——以芯片领域为例[J]. 中国市场, 2020(35): 32-35.
39 Wang X, He X N, Wang M, et al. Neural graph collaborative filtering[C]// Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM Press, 2019: 165-174.
40 He X N, Deng K, Wang X, et al. LightGCN: simplifying and powering graph convolution network for recommendation[C]// Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York: ACM Press, 2020: 639-648.
41 黄春淦, 王桂平, 吴波, 等. 基于轻量级图卷积和隐式反馈增强的多样化推荐[J]. 计算机科学, 2024, 51(S1): 681-691.
42 何喜军, 吴爽爽, 武玉英, 等. 基于属性异构网络表示学习的专利交易推荐[J]. 情报学报, 2022, 41(11): 1214-1228. |
|
|
|
