自适应时序图网络的科研合作推荐方法

doi:10.16088/j.issn.1001-6600.2026032301

摘要/Abstract

摘要： 癌症科研合作推荐可建模为连续时间动态图上的链路预测问题。现有时序图模型多面向通用场景,难以有效融合领域特定语义信息。为解决上述问题,本文提出一种面向科研合作推荐的自适应时序图网络模型AdaTGN-CR。该模型构建专家驱动的自适应邻居采样机制,根据专家学术活跃度、历史合作强度和专长水平动态调整邻居采样规模与邻居重要性权重;设计专家特征建模方法,对学术活跃度、合作强度及专长水平进行量化表示;引入层归一化GRU模块,以提升记忆更新稳定性并刻画长期合作模式。基于癌症领域数据集CATCD的实验结果显示,该模型在传导与归纳场景下的平均精度分别达到93.99%和93.81%,均优于代表性基线模型。实验结果表明,融合领域语义信息的自适应时序图建模方法能够有效提升科研合作推荐性能。

关键词: 动态图神经网络, 科研合作推荐, 专家特征, 时序图网络, 自适应邻居采样, 链路预测

Abstract: Cancer research collaboration recommendation can be naturally formulated as a link prediction problem on continuous-time dynamic graphs. However, existing temporal graph models are primarily designed for general scenarios and struggle to incorporate domain-specific semantic information. To address this limitation, an adaptive temporal graph network model, termed AdaTGN-CR, is proposed for research collaboration recommendation. Specifically, an expert-driven adaptive neighbor sampling mechanism is introduced to dynamically adjust the neighborhood size and neighbor importance weights according to expert status. In addition, an expert feature modeling method is developed to quantify academic activity, collaboration intensity, and expertise level. Furthermore, a layer-normalized gated recurrent unit (GRU) is employed to enhance the stability of temporal memory updates and capture long-term collaboration patterns. Experimental results on the cancer research dataset CATCD demonstrate that the proposed model achieves average precision (AP) of 93.99% and93.81% under transductive and inductive settings, respectively, consistently outperforming state-of-the-art baseline models. These results indicate that integrating domain semantic information with adaptive temporal graph modeling significantly enhances the performance of research collaboration recommendation.

Key words: dynamic graph neural networks, research collaboration recommendation, expert features, temporal graph networks, adaptive neighbor sampling, link prediction

中图分类号: TP391.1

王慧, 杨志成, 祝君豪, 胡昌志. 自适应时序图网络的科研合作推荐方法[J]. 广西师范大学学报（自然科学版）, 2026, 44(6): 69-86.

Wang Hui, Yang Zhicheng, Zhu Junhao, Hu Changzhi. Adaptive temporal graph network for research collaboration recommendation[J]. Journal of Guangxi Normal University(Natural Science Edition), 2026, 44(6): 69-86.

参考文献

[1] Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA: A Cancer Journal for Clinicians, 2024, 74(3): 229-263. DOI:10.3322/caac.21834.
[2] Stokols D, Hall K L, Taylor B K, et al. The science of team science: overview of the field and introduction to the supplement[J]. American Journal of Preventive Medicine, 2008, 35(2 Suppl): S77-S89. DOI: 10.1016/j.amepre.2008.05.002.
[3] 丁步月, 支凤稳, 郑彦宁. 科研合作意愿影响因素元分析[J]. 信息与管理研究, 2026, 11(1): 74-85. DOI:10.3969/j.issn.2096-2827.2026.01.008.
[4] 刘竟, 孙薇. 基于链路预测的潜在科研合作关系发现研究[J]. 情报理论与实践, 2017, 40(7): 88-92, 121. DOI: 10.16353/j.cnki.1000-7490.2017.07.016.
[5] 谢小东, 吴洁, 盛永祥, 等. 基于图卷积网络的发明人跨领域合作伙伴识别方法[J]. 情报杂志, 2024, 43(4): 175-183, 167. DOI:10.3969/j.issn.1002-1965.2024.04.023.
[6] Kumar S, Zhang X K, Leskovec J. Predicting dynamic embedding trajectory in temporal interaction networks[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York, NY: Association for Computing Machinery, 2019: 1269-1278. DOI: 10.1145/3292500.3330895.
[7] Trivedi R, Farajtabar M, Biswal P, et al. DyRep: learning representations over dynamic graphs[C]//International Conference on Learning Representations (ICLR 2019). Appleton, WI: ICLR, 2019: 1-25.
[8] Rossi E, Chamberlain B, Frasca F, et al. Temporal graph networks for deep learning on dynamic graphs[PP/OL]. V3. arXiv(2020-10-09)[2026-03-23].https://doi.org/10.48550/arXiv.2006.10637.
[9] 刘琳岚, 冯振兴, 舒坚. 基于时序图卷积的动态网络链路预测[J]. 计算机研究与发展, 2024, 61(2): 518-528. DOI: 10.7544/issn1000-1239.202220776.
[10]李聪, 季新生, 刘树新, 等. 基于节点匹配度的动态网络链路预测方法[J]. 网络与信息安全学报, 2022, 8(4): 131-143. DOI: 10.11959/j.issn.2096-109x.2022053.
[11]Zhu J, Yaseen A. A recommender for research collaborators using graph neural networks[J]. Frontiers in Artificial Intelligence, 2022, 5: 881704. DOI: 10.3389/frai.2022.881704.
[12]Wang J H, Mei H Y, Li K, et al. Collaborative filtering model of graph neural network based on random walk[J]. Applied Sciences, 2023, 13(3): 1786. DOI: 10.3390/app13031786.
[13]王慧, 祝君豪, 杨志成, 等. 基于多语义与时序卷积的知识图谱推荐模型[J]. 广西师范大学学报(自然科学版), 2026, 44(3): 107-120. DOI: 10.16088/j.issn.1001-6600.2025071802.
[14]程章桃, 钟婷, 张晟铭, 等. 基于图学习的推荐系统研究综述[J]. 计算机科学, 2022, 49(9): 1-13. DOI: 10.11896/jsjkx.210900072.
[15]于蒙, 何文涛, 周绪川, 等. 推荐系统综述[J]. 计算机应用, 2022, 42(6): 1898-1913. DOI: 10.11772/j.issn.1001-9081.2021040607.
[16]吴静, 谢辉, 姜火文. 图神经网络推荐系统综述[J]. 计算机科学与探索, 2022, 16(10): 2249-2263. DOI: 10.3778/j.issn.1673-9418.2203004.
[17]马帅, 刘建伟, 左信. 图神经网络综述[J]. 计算机研究与发展, 2022, 59(1): 47-80. DOI: 10.7544/issn1000-1239.20201055.
[18]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, NY: Association for Computing Machinery, 2020: 639-648. DOI: 10.1145/3397271.3401063.
[19]Wang X, He X N, Cao Y X, et al. KGAT: knowledge graph attention network for recommendation[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York, NY: Association for Computing Machinery, 2019: 950-958. DOI: 10.1145/3292500.3330989.
[20]Fu X Y, Zhang J N, Meng Z Q, et al. MAGNN: metapath aggregated graph neural network for heterogeneous graph embedding[C]//Proceedings of The Web Conference 2020. New York, NY: Association for Computing Machinery, 2020: 2331-2341. DOI: 10.1145/3366423.3380297.
[21]Gilmer J, Schoenholz S S, Riley P F, et al. Neural message passing for quantum chemistry[C]//Proceedings of the 34th International Conference on Machine Learning (PMLR 70). Cambridge, MA: JMLR, 2017: 1263-1272.
[22]Kipf T N, Welling M. Semi-supervised classification with graph convolutional networks[C]//International Conference on Learning Representations (ICLR 2017). Appleton, WI: ICLR, 2017: 1-14.
[23]Veličković P, Cucurull G, Casanova A, et al. Graph attention networks[C]//International Conference on Learning Representations (ICLR 2018). Appleton, WI: ICLR, 2018: 1-12.
[24]Hamilton W L, Ying R, Leskovec J. Inductive representation learning on large graphs[C]//Advances in Neural Information Processing Systems 30 (NIPS 2017). Red Hook, NY: Curran Associates Inc., 2017: 1025-1035.
[25]Feng Y F, You H X, Zhang Z Z, et al. Hypergraph neural networks[J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2019, 33(1): 3558-3565. DOI: 10.1609/aaai.v33i01.33013558.
[26]朱宇航, 刘树新, 吉立新, 等. 一种融合局部拓扑影响力的时序链路预测算法[J]. 电子与信息学报, 2022, 44(4): 1440-1452. DOI:10.11999/JEIT210019.
[27]Xu D, Ruan C W, Korpeoglu E, et al. Inductive representation learning on temporal graphs[C]//International Conference on Learning Representations (ICLR 2020). Appleton, WI: ICLR, 2020: 1-19.
[28]Wang Y B, Chang Y Y, Liu Y Y, et al. Inductive representation learning in temporal networks via causal anonymous walks[C]//International Conference on Learning Representations (ICLR 2021). Appleton, WI: ICLR, 2021: 1-21.
[29]Yu Z Y, Wu J Q, Wu Z H, et al. TAMI: taming heterogeneity in temporal interactions for temporal graph link prediction[C]//Advances in Neural Information Processing Systems 38 (NeurIPS 2025). Red Hork, NY: Curran Associates Inc., 2025: 130380-130414. DOI: 10.52202/085713-3923.
[30]Li H Y, Xu Y M, Li Y M, et al. When speed meets accuracy: an efficient and effective graph model for temporal link prediction[J]. Proceedings of the VLDB Endowment, 2025, 18(10): 3396-3405. DOI: 10.14778/3748191.3748203.
[31]Dey R, Salem F M. Gate-variants of gated recurrent unit (GRU) neural networks[C]//2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS). Piscataway, NJ: IEEE, 2017: 1597-1600. DOI: 10.1109/MWSCAS.2017.8053243.
[32]杨兴耀,刘岩松,于炯,等.融合用户兴趣边界与可学习滤波器的序列化推荐模型[J].东北师大学报(自然科学版), 2025, 57(2):82-89. DOI:10.16163/j.cnki.dslkxb202209270001.
[33]张元钧, 张曦煌. 基于图卷积与长短期记忆网络的动态网络表示学习模型[J]. 计算机应用, 2021, 41(7): 1857-1864. DOI: 10.11772/j.issn.1001-9081.2020081304.
[34]Ba J L, Kiros J R, Hinton G E. Layer normalization[PP/OL]. arXiv(2016-07-21)[2026-03-23]. https://doi.org/10.48550/arXiv.1607.06450.
[35]Rendle S, Freudenthaler C, Gantner Z, et al. BPR: Bayesian personalized ranking from implicit feedback[C]//Proceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence. Arlington, TX: AUAI Press, 2009: 452-461.
[36]Cong W L, Zhang S, Kang J, et al. Do we really need complicated model architectures for temporal networks?[C]//International Conference on Learning Representations (ICLR 2023). Appleton, WI: ICLR, 2023: 1-27.
[37]Kang H, Ho J H, Mesquita D, et al. Online graph nets[C]//International Conference on Learning Representations (ICLR 2022). Appleton, WI: ICLR, 2022: 1-19.
[38]Souza A H, Mesquita D, Kaski S, et al. Provably expressive temporal graph networks[C]//Advances in Neural Information Processing Systems 35 (NeurIPS 2022). Red Hork, NY: Curran Associates Inc., 2022: 32257-32269. DOI: 10.52202/068431-2337.
[39]Zhang S W, Chen X, Xiong Y, et al. Towards adaptive neighborhood for advancing temporal interaction graph modeling[C]//KDD’24: Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York, NY: Association for Computing Machinery, 2024: 4290-4301. DOI: 10.1145/3637528.3671877.
[40]Tian Y X, Qi Y Y, Guo F. FreeDyG: frequency enhanced continuous-time dynamic graph model for link prediction[C]//International Conference on Learning Representations (ICLR 2024). Appleton, WI: ICLR, 2024: 1-20.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed