Research on Text Classification Method Based on Multi-type Classifier Fusion

DOI

10.2991/snce-18.2018.163How to use a DOI?

Keywords

Text classification; Classifier fusion; Principal component analysis; Latent semantic index

Abstract

Most of the traditional text classification methods use a single classifier, but different classifiers have different emphasis on classification tasks, which makes a single classification method have some limitations. This paper presents a text classification method based on multi-type classifier fusion, which uses word2vecand principal component analysis (PCA) as feature extraction method for multi-type classifier fusion. At the same time, the problem of category information is ignored in the weighted voting method of multi-type classifier, and the method of classifier weight calculation is adopted. The experimental results show that the multi-type classifier fusion method is in binary. Good performance has been obtained on corpus, multivariate corpus and specific corpus. The classifier weight calculation method with class weighting improves the classification performance by 1.19% compared with multi-type classifier fusion method.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Meilin Zeng
PY  - 2018/05
DA  - 2018/05
TI  - Research on Text Classification Method Based on Multi-type Classifier Fusion
BT  - Proceedings of the 8th International Conference on Social Network, Communication and Education (SNCE 2018)
PB  - Atlantis Press
SP  - 798
EP  - 805
SN  - 2352-538X
UR  - https://doi.org/10.2991/snce-18.2018.163
DO  - 10.2991/snce-18.2018.163
ID  - Zeng2018/05
ER  -