Attribute network semi-supervised community discovery method based on non-negative matrix three-factor decomposition

Abstract

The invention discloses an attribute network semi-supervised community discovery method based on non-negative matrix three-factor decomposition, which comprises the following steps of: firstly, constructing a matrix decomposition model combining links and contents, including three parts of topology, prior and content matrixes, and depicting the meaning of each variable in the model in detail; then, optimizing the model, solving partial derivatives of unknown variables in the model, and setting the partial derivatives to be 0 to obtain an updating rule of each variable; collecting and processing data, and extracting a required adjacency matrix, priori information and a content matrix from the attribute network; and randomly initializing parameters and unknown variables thereof, performing atraining process by using the updating rule about each unknown variable obtained in the step (2) by adopting a gradient descent method, putting a processed data set into the model in the step (1) fortraining, and continuously performing iterating until the parameters are updated and converged.

Description

technical field [0001] The invention belongs to the fields of machine learning, complex network and natural language processing, mainly relates to the fusion of network information, and proposes a method for dimensionality reduction and information fusion by using a non-negative matrix decomposition technology, and specifically relates to a non-negative matrix-based three-factor decomposition method. The method of semi-supervised community discovery in attribute networks is widely used in community discovery in attribute networks. Background technique [0002] With the development of the Internet, online social networks generate more and more data, both links and semantic content, such as user blogs, research papers, etc. This data is often modeled as a network of attributes, with links forming the topology of the graph and content modeled as attributes of the nodes in the graph. It is of great significance to discover the semantic communities of these networks. For exampl...

AI Technical Summary

Problems solved by technology

However, when the network is very sparse and the community structure is too vague, these methods often cannot accurately identify the community structure and its semantic information due to the singleness of information.

Images