Title: Chinese text classification based on character-level CNN and SVM

Authors: Huaiguang Wu; Daiyi Li; Ming Cheng

Addresses: School of Computer and Communication Engineering, Zhengzhou University of Light Industry, Zhengzhou, China ' School of Computer and Communication Engineering, Zhengzhou University of Light Industry, Zhengzhou, China ' The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China

Abstract: Aiming at the problems of curse of dimensionality, sparse data and long computation time in traditional SVM classification algorithm based on term frequency-inverse document frequency (TF-IDF), we propose a novel hybrid system for Chinese text classification: CSVM, which is independent of the hand-designed features and domain knowledge. Firstly, the encoding words are done by constructing a text vocabulary of size m for the input language, and then quantise each word using 1-of-m encoding. Secondly, we exploit the convolutional neural network (CNN) to extract the morphological features of character vectors from each word, and then through large scale text material training the semantic feature of each word vectors are be obtained the semantic feature of each word vectors. Finally, the text classification is carried out with the SVM multiple classifier. The experimental results show that the CSVM algorithm is more effective than other traditional Chinese text classification algorithm.

Keywords: term frequency-inverse document frequency; TF-IDF; support vector machine; SVM; character-level CNN; text vectorisation; text classification; Chinese text classification; character-level CNN; TF-IDF; extract features.

DOI: 10.1504/IJIIDS.2019.102940

International Journal of Intelligent Information and Database Systems, 2019 Vol.12 No.3, pp.212 - 228

Received: 13 Sep 2018
Accepted: 28 Feb 2019
Published online: 11 Oct 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article