Subjectivity Analysis of Code-Mixed (English-Chinese) Text Translated to English Using Supervised Learning

Law, Jian Wen (2020) Subjectivity Analysis of Code-Mixed (English-Chinese) Text Translated to English Using Supervised Learning. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Law Jian Wen_Fulltext.pdf Restricted to Registered users only Download (1MB)

Abstract

Chinese speaking Malaysian social media users on Facebook and Twitter has the behavior of using Chinese words in the comments of a post that is contain mainly in English. Preprocessing will be performed to reduce the noises of the text and then translation of the code-mixed Chinese-English to a standard language format which is English language is performed. English language will be the translated text as used to perform subjectivity analysis because most of the libraries and models are having the well performance and accuracy on English language. Supervised learning is used to perform subjectivity analysis on the translated text. The proposed system will get the important inputs from the comments made by the author who comment on the post in either format of CSV file or string. After subjectivity analysis has been performed on the input text, the result would be presented to the system users which are typically organizations interested to determine the sentiments of the public towards their products and brands. They can use these results to understand the feedback given by their customers and make decisions. Next, the methodology that would apply in this project is waterfall model where breakdown the project activities into a sequential phase and perform it phase by phase.