Skip to content

Latest commit

 

History

History
30 lines (19 loc) · 1.18 KB

README.md

File metadata and controls

30 lines (19 loc) · 1.18 KB
Raw

Parliamanet Election NLP

About

NLP Analyse of election machine texts from the Finnish parliament elections 2015.

The data used for the prediction comes from avoindata.fi: https://www.avoindata.fi/data/fi/dataset/eduskuntavaalien-2015-ylen-vaalikoneen-vastaukset-ja-ehdokkaiden-taustatiedot

The result of the analysis is that the model could make party predictions of any given candidates "I want to work towards"- statemanet with approx. 44% accuracy on the test data. A complete report and statistics can be found under reports: https://github.com/chpatola/election_nlp/tree/master/reports

How to use

  1. Clone this repository to your own computer
  2. cd to the repository
  3. Create a new conda environment where is the name you want to give the new environment. Then activate the enivronment
conda create --name <env> --file requirement.txt
  
conda activate <env>

  1. Set your base path at line 7 in predict_model.py

  2. Run the code from the file predict_model.py.

python predict_model.py
  1. The programme will print some output of the analysis but you will also find results in the reports folder: https://github.com/chpatola/election_nlp/tree/master/reports