Comparing the voting pattern of citizens to their Congressional Representative
- Scrape the Congress.gov listing of role call votes in the US House of Representatives and the Senate between now and 2008 (subject to change)
- Grab the voting history of Congressional districts for the same period of time from the OpenElections project.
- Map the Congressional Districts in the OpenElections data to the representatives in the Congrress.gov data
- Use the Google Natural Language API to perform sentiment analysis and categorization on the text of ballot measures for both citizens and representatives
- Model representatives and Congressional districts as an aggregate of their voting for or against those sentiments
- Evaluate the similarity between Congressional district and its representative
- 2-7-2018 At the suggestion of Dr. Hearne, looking to find some subset of census data to use to build up a profile of citizens in a congressional district.
Inserted Representatives into the Representatives MySQL table.
-
2-3-2018 Reviewed the OpenElections data, and discovered that the amount of data on ballot measures would be insufficient to perform a large scale analysis on the sentiments of voting precincts based on sentiment analysis of those ballot measure. We resolve to perform that same sentiment analysis on the the text of matters before Congress to create a profile of each Congressperson, and further use Census data to create a profile of that Congressperson's constituents. We will use these profiles to perform outlier detection on the interaction between Congresspeople and their constituents.
-
1-27-2018 Initial Commit