Presidential Job Approval Rating Analysis Through Social Media

Author

Subramanian VenkataramanFollow
Subramanian Venkataraman, Harrisburg University of Science and TechnologyFollow

Completion Date

Fall 10-15-2017

Document Type

Dissertation

Degree Name

Master of Science (MS)

Program or Discipline Name

Analytics

First Advisor

Professor Dr. Rand Ford

Second Advisor

Professor Phil Grim

Third Advisor

Mark Newman

Abstract

The aim of this study is to identify patterns in President Trump’s approval in the

Twitter universe through Social Media and Sentiment Analysis, and compare

against scientific polling to get meaningful insights on the limitations of Social

Media Analytics. For the purposes for this exercise, results from scientific polling

will be considered the true measure of approval, and will be used as control. In

order to perform sentiment analysis, we have used supervisory learning using

Naive Bayes Classifier algorithm which produced 0.862667 accuracy levels.

Comments

The Naïve Bayes classifier has produced sentiment results almost equivalent to

President job approval rating company’s results. The trends and spikes in the

graph for both approval rating point and sentiment factors are the reflection of

some of the major events, criticism that happened in the last 3 weeks of time

27

which are related to Hurricane Maria, attack on North Korea and tax reforms. The

trend between approval rating points and Twitter Sentiments are almost trending

in line with each other.

The Twitter Sentiment results are almost trending with some of the key Approval

Rating Company’s reports such as Rasmussen Reports, Gallup Reports, Morning

Consult and ISPOS for Approval and Disapproval ratings This indicates that there

is a scope to understand President Job approval rating in real time using Twitter

Sentiments. Also, there is a possibility for the President and Republicans to

influence the approval rating results by sending right signals to the public through

social media.

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Recommended Citation

Venkataraman, S., & Venkataraman, S. (2017). Presidential Job Approval Rating Analysis Through Social Media. Retrieved from https://digitalcommons.harrisburgu.edu/anms_dandt/5