Title: Personality modelling and sentiment analysis on Chinese micro-blog posts

Authors: Kai Gao; Chongyang Yue; Siyu Li; Duoxing Liu; Erliang Zhou; Herbert Daly

Addresses: School of Information Science and Engineering, Hebei University of Science and Technology, No. 26 Yuxiang Street, Shijiazhuang Hebei, 050000, China; Department of Computer Science and Technology, University of Bedfordshire, Luton, Bedfordshire, LU1 3JU, UK ' School of Information Science and Engineering, Hebei University of Science and Technology, No. 26 Yuxiang Street, Shijiazhuang Hebei, 050000, China; Department of Computer Science and Technology, University of Bedfordshire, Luton, Bedfordshire, LU1 3JU, UK ' School of Information Science and Engineering, Hebei University of Science and Technology, No. 26 Yuxiang Street, Shijiazhuang Hebei, 050000, China; Department of Computer Science and Technology, University of Bedfordshire, Luton, Bedfordshire, LU1 3JU, UK ' School of Information Science and Engineering, Hebei University of Science and Technology, No. 26 Yuxiang Street, Shijiazhuang Hebei, 050000, China; Department of Computer Science and Technology, University of Bedfordshire, Luton, Bedfordshire, LU1 3JU, UK ' School of Information Science and Engineering, Hebei University of Science and Technology, No. 26 Yuxiang Street, Shijiazhuang Hebei, 050000, China; Department of Computer Science and Technology, University of Bedfordshire, Luton, Bedfordshire, LU1 3JU, UK ' School of Information Science and Engineering, Hebei University of Science and Technology, No. 26 Yuxiang Street, Shijiazhuang Hebei, 050000, China; Department of Computer Science and Technology, University of Bedfordshire, Luton, Bedfordshire, LU1 3JU, UK

Abstract: Intelligent information process such as opinion mining and sentiment analysis on social media remains an ongoing challenge, and it is also useful in public opinion surveillance. Analysing micro-blog posts is often hampered by their very brief content as well as the use of misspelled or abbreviated words. This paper focuses on personality modelling and sentiment analysis on Chinese micro-blog posts. Social media data pre-processing, to identify named entities and word sense disambiguation, is essential. The proposed pre-processing includes the double-array trie based segmentation and viterbi based word sense disambiguation, together with the co-occurrence probability based processing of unknown words. The personality modelling procedure vectorises micro-blog posts into high dimension eigenvectors. As for the sentiment analysis, this paper proposes the multi-convolutional neural method to solve the sentiment tendency determination problem. The experimental results show the feasibility of the approach, and existing problems and future works are also present in the end.

Keywords: intelligent information; opinion mining; sentiment analysis; social media; micro-blog; personality modelling; pre-processing; double-array trie; segmentation; word sense disambiguation; neural; tendency.

DOI: 10.1504/IJIIDS.2018.091617

International Journal of Intelligent Information and Database Systems, 2018 Vol.11 No.1, pp.67 - 78

Accepted: 15 Oct 2017
Published online: 08 May 2018 *

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