Universiti Teknologi Malaysia Institutional Repository

Chinese sentence similarity calculation based on modifiers

Wang, Fangling and Ye, Shaoqiang and Kang, Diwen and Mohd. Zain, Azlan and Zhou, Kaiqing (2022) Chinese sentence similarity calculation based on modifiers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 13338 (NA). pp. 301-310. ISSN 0302-9743

Full text not available from this repository.

Official URL: http://dx.doi.org/10.1007/978-3-031-06794-5_25

Abstract

To compute the similarity of Chinese sentences accurately, a revised Chinese sentence similarity approach is proposed though enhancing the importance of the modifiers of stem of sentence. After extracting the modified part of the sentence by Language Technology Platform (LTP), this part of each structure could be removed the longest common substring, to better capture the similarities of modified parts. The entire method includes three phases, which are to split the sentences into principal and predicate object structures using the syntactic analysis tool, to generate modifiers and sentence stem vectors and calculate the similarity between the vectors using the Word2Vec, and to obtain the similarity between two sentences by weighting each part. Experimental results on 200 sentences of the LCQMC dataset and corresponding analysis reveal that the proposed method can obtain more accurate similarity calculation results by effectively gaining the modified part - which affects the whole sentence meaning effectively-of the sentence structure.

Item Type:Article
Uncontrolled Keywords:Chinese sentence similarity, Natural language processing, Syntactic structure, Word vector, Word2Vec
Subjects:Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:Computing
ID Code:100496
Deposited By: Widya Wahid
Deposited On:14 Apr 2023 02:15
Last Modified:14 Apr 2023 02:15

Repository Staff Only: item control page