Journal of Computer Applications ›› 2024, Vol. 44 ›› Issue (9): 2726-2731.DOI: 10.11772/j.issn.1001-9081.2023091325

• Data science and technology • Previous Articles     Next Articles

Constructing pre-trained dynamic graph neural network to predict disappearance of academic cooperation behavior

Yu DU, Yan ZHU()   

  1. School of Computing and Artificial Intelligence,Southwest Jiaotong University,Chengdu Sichuan 611756,China
  • Received:2023-09-27 Revised:2023-12-11 Accepted:2023-12-12 Online:2023-12-22 Published:2024-09-10
  • Contact: Yan ZHU
  • About author:DU Yu, born in 1999, M. S. candidate. Her research interests include network analysis, academic network cooperative behavior disappearance prediction.
  • Supported by:
    Science and Technology Program of Sichuan Province(2019YFSY0032)

构建预训练动态图神经网络预测学术合作行为消失

杜郁, 朱焱()   

  1. 西南交通大学 计算机与人工智能学院,成都 611756
  • 通讯作者: 朱焱
  • 作者简介:杜郁(1999—),女,四川雅安人,硕士研究生,主要研究方向:网络分析、学术网络合作行为消失预测
    朱焱(1965—),女,广西桂林人,教授,博士,CCF会员,主要研究方向:社交网络计算、大数据、数据挖掘。
  • 基金资助:
    四川省科技计划项目(2019YFSY0032)

Abstract:

Some of the existing research on link disappearance only focuses on discovering and analyzing the reasons for link disappearance, while some research only uses static network representations for prediction, rarely analyzing link disappearance problem from the perspective of network dynamic evolution. In response to the above research status, a pre-trained dynamic graph neural network based academic cooperative behavior disappearance prediction model PreDGN (Pre-trained Dynamic Graph neural Network) was proposed. In PreDGN, firstly, the temporal information of the dynamic network was captured by using the dynamic graph to generate pre-training tasks, and the topological information of the network was supplemented with edge features constructed by temporal motifs. Then, combined with attention node embedding based on time encoding, node representations were learnt more accurately. The historical information of the dynamic graph was learnt by the pre-trained model and the model was able to be fine-tuned in specific tasks for predicting the disappearance of academic cooperation behaviors. Experiments were conducted on data from the publicly available academic cooperation dataset HepTh with different time spans and data scales. On the 1996, 1997, 94—96, and 97—99 subsets, compared to the second best method: dynamic graph neural network DyRep, the proposed model has the Area Under the ROC Curve (AUC) increased by 10.47, 8.16, 13.41, and 3.27 percentage points, respectively, and the Average Precision (AP) improved by 5.87, 2.15, 8.26, and 3.01 percentage points, respectively.

Key words: dynamic graph, pre-training, graph representation learning, graph neural network, network motif

摘要:

现有链接消失问题研究工作一部分只停留在发现和分析链接消失的原因上,一部分仅使用静态网络表示进行预测,很少从网络动态演化的角度分析链接消失预测问题。针对以上研究现状,提出一种预训练动态图神经网络学术合作行为消失预测模型PreDGN(Pre-trained Dynamic Graph neural Network)。PreDGN首先通过动态图生成预训练任务捕捉动态网络的时间信息,同时利用时序模体构造的边特征补充网络的拓扑信息;其次,结合基于时间编码的注意力节点嵌入,能够更精准地学习节点的表征。经过预训练的模型学习了动态图的历史信息,而且可以在特定的学术合作行为消失预测任务中进行微调。使用公开学术合作数据集HepTh中不同时间跨度、不同数据规模数据进行实验的结果表明,在1996、1997、94—96和97—99子数据集上,相较于次优的动态图神经网络方法(DyRep),所提模型的曲线下面积(AUC)指标分别提高了10.47、8.16、13.41和3.27个百分点,平均精度(AP)指标分别提高了5.87、2.15、8.26和3.01个百分点。

关键词: 动态图, 预训练, 图表示学习, 图神经网络, 网络模体

CLC Number: