An Improved Method of Computing Chinese Sentence Similarity

DOI

10.2991/icicci-15.2015.7How to use a DOI?

Keywords

Keywords-Sentence Similarity; HowNet; CiLin; Tree Kernel; Maximum Weight Matching

Abstract

Abstract—The Chinese sentence similarity has been used widely in the field of Chinese information processing. There are many methods proposed to measure the similarity of Chinese sentences, but the accuracy of these methods is still lower. In order to improve this problem, this paper proposes an improved method to compute the similarity of Chinese sentences based on predecessors' research. The method proposed not only takes into account semantic meaning of sentences based on HowNet, CiLin and Maximum Weight Matching, but also considers the syntactic information based on tree kernel algorithm. Meanwhile, some special cases, such as antonym, double negation, are also considered. Experiments show that this improved method has relatively higher accuracy in Chinese sentence similarity computation.

Copyright

© 2015, 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  - Lu Wang
AU  - Zhongshi He
PY  - 2015/09
DA  - 2015/09
TI  - An Improved Method of Computing Chinese Sentence Similarity
BT  - Proceedings of the 2nd International Conference on Intelligent Computing and Cognitive Informatics
PB  - Atlantis Press
SP  - 29
EP  - 33
SN  - 1951-6851
UR  - https://doi.org/10.2991/icicci-15.2015.7
DO  - 10.2991/icicci-15.2015.7
ID  - Wang2015/09
ER  -