Ross Millward Undergraduate Dissertation 2015/16

The Twitterati: Text Analysis of Twitter and Other Social Media

Supervised by M.Hepple

Abstract

Millions of people log in everyday to micro- blogging  site Twitter to post short updates on their opinions and thoughts. These small messages from people around the world are of huge interest to companies and individuals as together they can provide a voice for a large portion  of the world's population. There are simply too many tweets posted every day for them to be manually processed and so Natural Language Processing can be called upon to automate the task. Furthering this, sentiment analysis can be used to allow a computer to automatically derive the sentiments demonstrated in a tweet and so, by careful selection of tweets, it is possible to estimate the sentiments of the general public towards a given  topic.

I will be investigating and experimenting with the possibility of using sentiment analysis applied to tweets to track and predict the movement of share values for various companies tradable through NASDAQ.