基于事件的异质信息网络表征学习

Event driven representation learning for heterogeneous information networks

傅国绩

11749236

硕士

计算机科学与技术

姚新

2019-05-30

2019-05-31

哈尔滨工业大学

深圳

近年来, 异质信息网络表征学习成为机器学习和数据挖掘领域的研究热点。然而现有的网络表征学习算法只考虑了网络中链接数量对节点相关性的影响，而忽略了链接本身的性质，从而会丢失原始网络中的一些语义信息。因此，本文所提出了一种基于事件的新的表征学习算法，能够更有效地保留原始异质信息网络的结构和语义信息。

Recently, heterogeneous information network representation learning has become one of the hot topics in machine learning and data mining communities. However, existing network representation learning methods consider only the impact of the quantities of relations in the network on the relevance of nodes and ignore their properties and hence they will lose some semantic information of a network. To tackle this issue, this paper proposes a new event-driven heterogeneous information network representation learning algorithm, which is more effective to preserve both structural and semantic information of a heterogeneous information network.

异质信息网络 表征学习 网络分析

heterogeneous information networks representation learning network analysis

中文

联合培养

南方科技大学

傅国绩. 基于事件的异质信息网络表征学习[D]. 深圳. 哈尔滨工业大学,2019.