A network representation learning method and device based on random walk of edges

Abstract

The invention discloses a network representation learning method and device based on random walk of edges, and the method comprises the steps: calculating the similarity between the edges in a networkaccording to a topic vector of each node of the network and an associated timestamp of the edge; Calculating the edge-to-edge transition probability according to the calculated edge-to-edge similarity; Based on the guidance of the meta-path, carrying out random walk according to the calculated transition probability to generate a node sequence; And performing representation learning of the node according to the obtained node sequence to obtain low-dimensional representation of the node. According to the method, semantic information and time information can be interpreted to obtain richer network content, so that potential information of a real world can be more truly and effectively mined; And the method can be used for carrying out more appropriate and practical representation on the network which changes along with the passage of time.

Description

technical field [0001] The present invention relates to the technical field of network representation learning, in particular to a network representation learning method and device based on edge random walk. Background technique [0002] Many applications in real life can be abstracted as networks, and networks can be represented by graphs. Therefore, most studies use some methods of graph research to help analyze networks, so as to solve the needs and problems of various practical scenarios. Graphs are an important data representation and are widely used in related fields such as computer science and biology. Real-life applications such as social networks, road networks, academic networks, biological protein networks, and communication networks can all be modeled as graphs. By modeling the interaction behavior between entities as a graph, researchers are able to understand various networks in a systematic way. Effective graph analysis can give users a deeper understanding...

AI Technical Summary

Problems solved by technology

[0004] Early network representation learning was mainly aimed at static homogeneous networks, which can effectively represent the network topology, but lacks detailed semantic information; moreover, the amount of network data is increasing day by day, and the scale and shape of the network change over time. There are significant changes, and the existing network representation learning mainly for static isomorphic networks also lacks time information, and cannot properly adjust the network representation for changes in the network due to the passage of time; therefore, the existing network Representation learning methods cannot interpret the rich content contained in the network, and cannot perform more appropriate and realistic representations for networks that change over time

Images