A Text Classification Method Based on Graph Kernel and Convolutional Neural Network

Abstract

The invention relates to a text classification method based on a graph kernel and a convolutional neural network, belonging to the technical fields of data mining and information retrieval. The core idea is: firstly, the text is preprocessed into a graph structure representation, in which the nodes in the graph correspond to the words in the text; then the weights of the nodes are calculated based on the graph structure, and then the graph structure is decomposed into multiple graph structures using the community discovery algorithm. Then use graph kernel technology to map the graph to a high-dimensional space to obtain the tensor representation of the graph, and finally input the tensor representation to the convolutional neural network to deeply mine the graph features and output the text category. Compared with the prior art, the present invention can make full use of the internal structure and context semantics of the text, so that the text content can be fully expressed; the node information is more reasonable; and the complex and cumbersome processing process in the text classification is effectively solved.

Description

technical field [0001] The invention relates to a text classification method, in particular to a text classification method based on a graph kernel and a convolutional neural network, and belongs to the technical fields of data mining and information retrieval. Background technique [0002] With the advent of the era of big data, the amount of information has exploded, and the way of information processing has gradually transitioned from traditional manual processing to automated processing. As an important task of information processing, text classification aims to automatically classify unlabeled documents into a predetermined category set, which can solve the phenomenon of information clutter to a large extent, and then realize efficient management of massive information. Text classification technology has been It is widely used in information filtering, information retrieval, topic detection and tracking and other fields. [0003] There are three main types of text clas...

AI Technical Summary

Problems solved by technology

[0007] For the problem of text classification, the existing technology mainly expresses the text as a vector space model, which loses the semantic structure information of the text. The present invention proposes a text classification method based on graph kernel and convolutional neural network, which can Effectively preserve the semantic structure of the text and improve classification accuracy

Images