Sentiment Analysis of Donald Trump's Tweets using Machine Learning

Document Type

Oral Presentation

Campus where you would like to present

Ellensburg

Event Website

https://digitalcommons.cwu.edu/source

Start Date

18-5-2020

Abstract

With social media websites growing in popularity every day, there is much untapped potential in leveraging this public information to learn many new things. Twitter alone has a huge user base of over 330 million active monthly users. This huge volume of tweets coupled with the informal nature of text makes processing this information very difficult. In recent years, the combination of Natural Language Processing (NLP) with Machine Learning (ML) has become increasingly prominent. The understanding of natural language by machines is still an open and challenging task in spite of the huge recent progress in man-machine interaction using deep learning. Twitter is a great choice for ML analysis due to its textbased nature and limited tweet length. For example, one could analyze Twitter for early reporting on big events that have not been reported or hit the news cycle yet. We apply state-of-the-art ML and NLP tools to Donald Trump tweets, with the goal of analyzing their sentiment. This approach, called Sentiment Analysis, attempts to determine if a text contains a positive or a negative sentiment. A clearly positive sentiment is something like “I love cake”, while a clearly negative sentiment is something like “I hate mosquito’s”. While these examples are obvious, that is not always the case. Analyzing Donald Trump’s tweets turns out to be a very complex task. As a result, our computer model can determine with a reasonable accuracy if a Donald Trump tweet is positive or negative.

Faculty Mentor(s)

Razvan Andonie

Department/Program

Computer Sciences

Additional Mentoring Department

https://cwu.studentopportunitycenter.com/2020/04/sentiment-analysis-of-donald-trumps-tweets-using-machine-learning/

Share

COinS
 
May 18th, 12:00 PM

Sentiment Analysis of Donald Trump's Tweets using Machine Learning

Ellensburg

With social media websites growing in popularity every day, there is much untapped potential in leveraging this public information to learn many new things. Twitter alone has a huge user base of over 330 million active monthly users. This huge volume of tweets coupled with the informal nature of text makes processing this information very difficult. In recent years, the combination of Natural Language Processing (NLP) with Machine Learning (ML) has become increasingly prominent. The understanding of natural language by machines is still an open and challenging task in spite of the huge recent progress in man-machine interaction using deep learning. Twitter is a great choice for ML analysis due to its textbased nature and limited tweet length. For example, one could analyze Twitter for early reporting on big events that have not been reported or hit the news cycle yet. We apply state-of-the-art ML and NLP tools to Donald Trump tweets, with the goal of analyzing their sentiment. This approach, called Sentiment Analysis, attempts to determine if a text contains a positive or a negative sentiment. A clearly positive sentiment is something like “I love cake”, while a clearly negative sentiment is something like “I hate mosquito’s”. While these examples are obvious, that is not always the case. Analyzing Donald Trump’s tweets turns out to be a very complex task. As a result, our computer model can determine with a reasonable accuracy if a Donald Trump tweet is positive or negative.

https://digitalcommons.cwu.edu/source/2020/COTS/46