Automated Sentiment Analysis
Abstract
This project develops a software tool to automate the collection of sentiment data from electronic media reports. Through analyses the report illustrates that the data generated from the new software tool produces valid and reliable data measuring attitudes and beliefs about political candidates, their actions, and their policies. The report further shows that such data improves the model fit of political behavior models. Specifically, the new data reveals important impacts on the actions and interactions of political actors. The report closes with the prospects for future analyses and research in this area and the area of national security.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2009
- Accession Number
- ADA532194
Entities
People
- Michael A. Covington
- Stephen M. Shellman
Organizations
- University of Georgia