Classification of user-level twitter polarity using soft computing approaches

• Main Author: Iyer, Shruthi Suresh
• Other Authors: Er Meng Joo
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/78280
• Institution: Nanyang Technological University

Sentiment Analysis belongs to a Machine Learning research area called Natural Language Processing or NLP. The purpose of NLP is to apply computational techniques and algorithms to help a computer analyze and understand natural language. Data, and more importantly knowledge obtained from data is invaluable. Enabled by the growing presence and influence of social media in everyday life, people tend to share increasing amounts of data about their lives online. By analyzing this data, it is possible to ‘mine’ the opinions or sentiments of the general public on any topic of interest. This is precisely what sentiment analysis aims to do. One such freely available source of online data is the popular social media platform – Twitter. Due of the brevity of tweets (limited to 280 characters) and the use of emojis, emoticons and acronyms, tweets are an especially unique source of a person’s opinions. This report covers the efforts and progress made in a yearlong project to develop a user-level sentiment classifier of tweets. It starts by examining the applicable computational approaches, followed by a relevant literature review of previous work in this area before diving into an analysis of the algorithmic approach, methodology, implementation, testing and findings of the project.