Vietnamese Sentence Similarity Based on Concepts

Abstract : We propose a novel method for measuring semantic similarity of two sentences. The originality of the method is the way that it explores the similarity of concepts referred to in the sentences using Wikipedia. The method also exploits Wiktionary to measure word-to-word similarity. The overall semantic similarity is a linear combination of word-to-word similarity, word-order similarity, and concept similarity. We build datasets consisting of 45 Vietnamese sentence pairs and then evaluate the method on these datasets. The results show that in the best cases, concept similarity help improving the performance of our method more than 15% point. The proposed method is language-independent and quite easy to employ. Therefore, one can readily adopt our method to measure semantic similarity for sentences written in other languages.
Liste complète des métadonnées

Cited literature [19 references]  Display  Hide  Download

https://hal.inria.fr/hal-01405592
Contributor : Hal Ifip <>
Submitted on : Wednesday, November 30, 2016 - 11:01:30 AM
Last modification on : Thursday, December 1, 2016 - 1:04:16 AM
Document(s) archivé(s) le : Monday, March 27, 2017 - 8:49:02 AM

File

978-3-662-45237-0_24_Chapter.p...
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

Citation

Hien Nguyen, Phuc Duong, Vinh Vo. Vietnamese Sentence Similarity Based on Concepts. 13th IFIP International Conference on Computer Information Systems and Industrial Management (CISIM), Nov 2014, Ho Chi Minh City, Vietnam. pp.243-253, ⟨10.1007/978-3-662-45237-0_24⟩. ⟨hal-01405592⟩

Share

Metrics

Record views

82

Files downloads

271