Title: A dynamic window split-based approach for extracting professional terms from Chinese courses

Authors: Shunxiang Zhang; Yanyong Du; Kui Lu

Addresses: School of Computer Science and Engineering, Anhui University of Science and Technology, Huainan, China ' School of Computer Science and Engineering, Anhui University of Science and Technology, Huainan, China ' School of Computer Science and Engineering, Anhui University of Science and Technology, Huainan, China

Abstract: Professional terms have an important role in representing knowledge and organisations of Chinese courses concisely and automatically. The biggest difficulty of extracting professional terms from Chinese courses is the uncertain length of professional terms. This paper proposes a dynamic window split-based approach for extracting professional terms from Chinese courses. First, content words in each Chinese course are achieved by utilising the ICTCLAS2013 system. Second, based on the achieved content words, two extracting algorithms of synthetic terms and non-synthetic terms are proposed. These two proposed extracting algorithms have embodied the idea of the dynamic window splitting. The experimental results show that the proposed approach can effectively extract professional terms from Chinese courses. Especially, the extracted synthetic terms keep the power-law characteristic of content words. Effective extracting of professional terms can provide the theory foundation for e-learning of Chinese course.

Keywords: semantic representation; content words; synthetic terms; window splitting; term extraction; China; professional terms; e-learning; electronic learning; online learning; professional courses; Chinese; computer science education.

DOI: 10.1504/IJCSE.2016.076948

International Journal of Computational Science and Engineering, 2016 Vol.12 No.4, pp.341 - 351

Received: 09 Jan 2015
Accepted: 22 Jan 2015
Published online: 08 Jun 2016 *

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