Title

Political Social Media Analysis Using Context-Dependent Artificial Intelligence

Presenter Information

Bradford RegeskiFollow

Faculty Mentor(s)

N/A

Campus

Dahlonega

Proposal Type

Oral Presentation

Subject Area

Computer Science

Location

MPR 3

Start Date

22-3-2019 1:00 PM

End Date

22-3-2019 2:00 PM

Description/Abstract

Social media has become an embedded part of the political system of the modern world, as regular citizens and political leaders have access to a digital platform to voice their personal opinions on policies and news. The recent 2016 General Election in the United States featured heavy use of social media platforms, but also featured accusations of election interference using biased and targeted social media posts and ads. The purpose of this research project is a proof-of-concept digital tool used to analyze posts and use artificial intelligence to dissect and analyze said posts for political keywords and opinions. Using context dependent artificial intelligence, with assistance from the Natural Language Toolkit1, the tool should be able to recognize current political keywords, recognize the sentiment of the post and categorize the post to measure the political leanings of the user. This tool could be theoretical election targeting tool used by domestic or foreign adversaries, and remains a prominent issue in the upcoming 2020 elections, and this research project can hopefully be an example that nearly anyone with technical skills can index and identify moving partisan trends in times of political upheaval.

1: https://www.nltk.org/

Media Format

flash_audio

This document is currently not available here.

Share

COinS
 
Mar 22nd, 1:00 PM Mar 22nd, 2:00 PM

Political Social Media Analysis Using Context-Dependent Artificial Intelligence

MPR 3

Social media has become an embedded part of the political system of the modern world, as regular citizens and political leaders have access to a digital platform to voice their personal opinions on policies and news. The recent 2016 General Election in the United States featured heavy use of social media platforms, but also featured accusations of election interference using biased and targeted social media posts and ads. The purpose of this research project is a proof-of-concept digital tool used to analyze posts and use artificial intelligence to dissect and analyze said posts for political keywords and opinions. Using context dependent artificial intelligence, with assistance from the Natural Language Toolkit1, the tool should be able to recognize current political keywords, recognize the sentiment of the post and categorize the post to measure the political leanings of the user. This tool could be theoretical election targeting tool used by domestic or foreign adversaries, and remains a prominent issue in the upcoming 2020 elections, and this research project can hopefully be an example that nearly anyone with technical skills can index and identify moving partisan trends in times of political upheaval.

1: https://www.nltk.org/