Social network feature dynamic extraction method based on vector compression and reconstruction

Abstract

The invention discloses a social network feature dynamic extraction method based on vector compression and reconstruction. The method comprises the steps of (1) generating a training set; (2) constructing a deep semi-supervised auto-encoder network; (3) constructing a generative adversarial network; (4) training a network; and (5) completing dynamic feature extraction of the social network in thegenerative adversarial network. According to the method, the deep semi-supervised auto-encoder network is constructed and trained, high-order social network structure information can be better captured, and based on a dynamic feature extraction method of the generative adversarial network, the method has relatively short processing time and relatively high space utilization rate when being used for processing a large-scale social network.

Description

technical field [0001] The invention belongs to the technical field of physics, and further relates to a dynamic extraction method of social network features based on vector compression and reconstruction in the technical field of vector representation. The method for dynamically extracting social network features of the present invention maintains the network topology and time series evolution characteristics of the social network in the form of feature vectors, and uses the extracted features for network structure data mining tasks such as social relationship discovery and community relationship division. Background technique [0002] The dynamic extraction method of social network features is based on the characteristics of network topology, through the dynamic modeling of the network, realizes the low-dimensional vectorized expression of the network, discovers the evolution law of the network and dynamically extracts the social network features. Neural network technologi...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to address the deficiencies in the above-mentioned prior art, and propose a dynamic extraction method of social network features based on vector compression and reconstruction, which is used to solve the problem that existing feature extraction methods ignore the high-order information of social network features. The problem of poor accuracy of the method of extracting social network features, and the problem of low space utilization and information distortion due to large calculation scale

Images