A Review on Sentiment Analysis for Code-Mix Chinese and English Text on Social Media

Abstract

Social media is rich with opinions. Millions of people shared their thoughts on products, services and events on Social Media Sites (SMS). Digital marketers extract and analyse content from SMS so that they know how best to promote their products or services to potential buyers. Government can get feedback from citizens about policies they have implemented. Works here reviews numerous sentiment analysis research works that study code-mix posts and comments that were expressed in formal and informal languages with a code-mix of Chinese and English or English and Hindi. Research in code-mix English and Hindi sentiment analysis are reviewed to provide some insights for application in code-mix Chinese and English. Raw data collected will be pre-processed into structured representation. Works here will discuss sentiment analysis that adopts lexicon approach, machine learning and combination of both. Works here will highlight translation and non-translation approaches used to analyse code-mix text. Discussion about propose solution for further exploration is discussion in a section. Critical remarks and a concluding section will be presented at the end of the paper. Keywords: code-mix, machine learning, lexicon