Bilingual sentiment detection - Investigating impact of tweet translation

Vimala, B. and Kaur, W. (2016) Bilingual sentiment detection - Investigating impact of tweet translation. In: Seventh International Conference on the Applications Digital Information and Web Technologies (ICADIWT 2016), 28 - 31 March 2016, Taipei, Taiwan.

[img]
Preview
PDF (Full text)
0001.pdf

Download (2MB)

Abstract

This paper looks at how translating tweets from Malay to English may impact its sentiment score, and if at all the score can be improved in case a tweet is translated compared to just translating the keywords. Tweets written in Malay were translated using an online dictionary before proceeding for analysis. An online sentiment analysis tool, Twin word was used to perform the sentiment analysis on both translated and untranslated tweets. The results of the analysis showed translating tweets did not create a significant impact on the overall sentiment score; therefore translating the whole length of the tweet would not affect the accuracy score.

Item Type: Conference or Workshop Item (Paper)
Funders: UNSPECIFIED
Uncontrolled Keywords: Sentiment detection, Twitter, bilingual tweets, tweet translation
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
T Technology > T Technology (General)
Divisions: Faculty of Computer Science & Information Technology
Depositing User: Mr. Mohd Safri
Date Deposited: 08 Apr 2016 02:22
Last Modified: 08 Apr 2016 02:22
URI: http://eprints.um.edu.my/id/eprint/15865

Actions (login required)

View Item View Item