融合知识图谱和图注意力网络的引文推荐算法  

Citation recommendation algorithm fusing knowledge graph and graph attention network

作　　者：樊海玮[1] 鲁芯丝雨 张丽苗 安毅生[1] FAN Haiwei;LU Xinsiyu;ZHANG Limiao;AN Yisheng(School of Information Engineering,Chang’an University,Xi’an Shaanxi 710064,China)

机构地区：[1]长安大学信息工程学院,西安710064

出　　处：《计算机应用》2023年第8期2420-2425,共6页journal of Computer Applications

基　　金：陕西高等教育教学改革研究项目(21BY031);陕西省第二批新工科研究与实践项目(54);中央高校基本科研业务费专项(300103112403)。

摘　　要：针对传统协同过滤(CF)中的数据稀疏和冷启动问题,以及元路径、随机游走算法没有充分利用节点信息的问题,提出融合知识图谱和图注意力网络的引文推荐算法(C-KGAT)。首先,使用TransR算法将知识图谱信息映射为低维稠密向量,以获取节点的嵌入特征表示;其次,利用图注意力网络通过多通道融合机制聚合邻居节点信息以丰富目标节点的语义,并捕获节点间高阶连通性;接着,在不影响网络的深度或宽度的情况下,引入动态卷积层动态地聚合邻居节点信息以提升模型的表达能力;最后,通过预测层计算用户和引文的交互概率。在公开数据集AAN(ACL Anthology Network)和计算机科学文献库(DBLP)上的实验结果表明,所提算法的效果优于所有对比模型,所提算法的MRR(Mean Reciprocal Rank)相较于次优模型NNSelect分别提升了6.0和3.4个百分点,所提算法的精确率和召回率指标也有不同程度的提升,验证了算法的有效性。Aiming at problems of data sparseness and cold start in traditional Collaborative Filtering(CF) and problem that meta-path and random walk algorithms do not fully utilize node information,a Citation Recommendation Algorithm Fusing Knowledge Graph and Graph Attention Network(C-KGAT) was proposed.Firstly,knowledge graph information was mapped into low-dimensional dense vectors by using TransR algorithm to obtain embedded feature representation of the nodes.Secondly,through multi-channel fusion mechanism,graph attention network was used to aggregate neighbor node information to enrich semantics of target nodes and capture high-order connectivity between nodes.Thirdly,without affecting depth or width of network,dynamic convolutional layer was introduced to aggregate information of neighbor nodes dynamically to improve expression ability of the model.Finally,the interaction probabilities of users and citations were calculated through the prediction layer.Experimental results on public datasets AAN(ACL Anthology Network) and DataBase systems and Logic Programming(DBLP) show that the proposed algorithm performs better than all comparison models.The MRR(Mean Reciprocal Rank) of the proposed algorithm is increased by 6.0 and 3.4 percentage points respectively compared with that of the suboptimal model NNSelect,and the Precision and Recall indicators of the proposed algorithm also have different degrees of improvement,which verifies the effectiveness of the algorithm.

关 键 词：知识图谱 图注意力网络 引文推荐 动态卷积 聚合 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]