Social media/news analysis system (for financial market and public relations)

• Main Author: Wan, Yuan
• Other Authors: Shao Xuguang Michelle
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/68259
• Institution: Nanyang Technological University

In the era of information, the ability to extract insightful and meaningful information from the ocean of Internet is highly appreciated. This thesis will focus on online sentiment analysis using Twitter and other free sources. The project consists the research and improvement on commonly used sentimental analysis approaches. The analyzer developed and trained in this project demonstrate superior capability in extracting sentiment from the short text, which is the majority of data created by the general public on The Internet. Geographic and other information are analyzed to assist the understanding of sentiments. To provide the user with a good understanding and better experience, this project uses an interactive website as the user interface. Besides, an automated alerting function is implemented to achieve intelligent monitoring. At the end of this project, more than 10 million of data entries were recorded. Four major case studies were conducted to demonstrate the capability of the system: sentimental analysis of candidates in the presidential election, an all-in-one new portal, a sentimental map generated based on Twitter, and sentiment email alerts.