Object vector representation generation method and device, electronic equipment and storage medium

Abstract

The invention provides an object vector representation generation method and device, electronic equipment and a storage medium. The method comprises the steps of determining a target path taking a target node corresponding to a target object as an end point according to an association relationship between nodes corresponding to various types of objects; determining a traversal sequence of each node on the target path; traversing each node on the target path according to the traversing sequence, aggregating the vector representation of each neighbor node of the node for each traversed node, generating the vector representation of the node until the target node is traversed, aggregating the vector representation of each neighbor node of the target node, and generating the vector representation of the target node. According to the method, semantics among nodes at different positions in the meta-path are distinguished, and the vector representation of each node can be more accurate by aggregating neighbor information at the position of each node on the meta-path.

Description

technical field [0001] The embodiments of the present application relate to the technical field of data processing, and in particular to a method, device, electronic device, and storage medium for generating a vector representation of an object. Background technique [0002] There is only one kind of node and edge in traditional Homogeneous Graph data, so when constructing a graph neural network, all nodes share the same model parameters and feature spaces of the same dimension. However, there can be more than one kind of nodes and edges in heterogeneous graph data, so different types of nodes are allowed to have characteristics or attributes of different dimensions, making heterogeneous graphs widely used. [0003] In a heterogeneous graph network, nodes can be connected by various types of relations such as meta-paths. Given a meta-path, each node has many meta-path-based neighbors. How to distinguish the subtle differences between neighbors and select some informative n...

AI Technical Summary

Problems solved by technology

[0004] However, in related technologies, when obtaining a vector representation of a node in a heterogeneous graph network, a neural network is directly used to obtain the weights of all neighbor nodes of the current node through the attention model on the heterogeneous graph network, and then The vector representation of the current node is obtained according to the weights of these neighbor nodes, and the semantic difference between nodes at different positions on the meta-path is not distinguished

Since each node has many meta-path-based neighbors, the vector representations under different meta-paths are different, and the vector representation of nodes cannot be accurately obtained without distinguishing the semantic differences between nodes

Images