Riley Hesbacher, [email protected]
In this project, I use the corpus of German political speeches to analyze sentiment between different parties. I compare the sentiment based on different parties and compare how the sentiment of individual parties changed over time and how those changes could align with current events of the time.
Adrien Barbaresi. A corpus of German political speeches from the 21st century. 11th Language Resources and Evaluation Conference (LREC 2018), May 2018, Miyazaki, Japan. pp.792-797. ffhal01798703f
This corpus consists of 6,685 speeches by 71 speakers, spanning a time from 1984 to 2017 and amounting to about 13 million words. These speeches come from speakers who are a part of the German government. The speeches are split into which office the speaker was from: the chancellery, the presidency, the Bundestag or the Ministry of Foreign Affairs. The corpus includes the speech, the speaker, the date and the location where the speech was made.
- final_report.md is my final report write-up.
- GPS_Sentiment_Analysis.ipynb is my code that includes my data cleaning, processing, and analysis. Here is the NB Viewer version.
- data-samples/ is where the original data files are found.
- image_files/ contains graphs of my data and findings.
- LICENSE.md contains licensing information.
- project_plan.md was my initial project plan.
- progress_reprt.md contains progress logs throughout this semester.
My Guestbook can be found here.