SENTIMENT ANALYSIS ON PRODUCT TWEETS

• Main Author: Ira Iryani , Ira Iryani
• Format: Final Year Project
• Language: English
• Published: Universiti Teknologi Petronas 2013
• Online Access: http://utpedia.utp.edu.my/13548/1/IRA%20IRYANI_15187.pdf; http://utpedia.utp.edu.my/13548/

With the advancement of Web 2.0, the user could do more than just retrieve information from a static website. Better user-interface, software and storage facilities all in one place which is called the web browser. One of the main features of this Web 2.0 includes the social media. The hype of social media such as Twitter and Facebook have made people express their opinions and feelings more easily publicly. Everyone interprets the information they got differently. They have their own understanding and interpretation on how the information is. With the technology that is rapidly growing, we can use the information that the user is displaying on social media and make this as opportunity thus identifying the problems as soon as it occurs. Sentiment analysis is about finding subjective information and grouped it into polarity classification (positive, negative or neutral). One of the objectives of this project is, to automatically categorize data into either positive sentiment, negative sentiment or neutral sentiment based on the subjective data that is obtained from the social media. This system can be useful for companies who are interested to get the fastest way to obtain juicy and latest information from the social network. Another target user could be the institutions that are reputation conscious. Case-Base Reasoning (CBR) will be used in this project. CBR is done by looking at past situations to solve the possible same current issue. Large amount of data is hard to comprehend thus, machine learning techniques could be used to automate the tasks and also provide the predictions over that matter.