btvote

We are looking at roll call votes taken in the German Bundestag during the last election period from 2017 to 2021. The ultimate goal of the project is to predict the 'closest' party based on the voting behavior. To achieve this goal, there are two steps taken. At first, we will train a model based on the voting behavior of the members of parliament. This will allow insights to which extent the voting behavior and the belonging party of a member of parliament are correlated. If a visible correlation between those two variables can be found, we will further train a model based on selected votes in the second step. The goal is to create a tool that is able to predict the party to which the user is closest to with respect to approximately 15 questions (roll call votes) the user has to answer beforehand. In the end, our work should result in a "ML Wahl-O-Mat", so a decision support for federal election using machine learning.