With the US election in 2008, social media became a central aspect of political campaigns. Campaigns analyze patterns, profiles, and optimal communication channels of the target audience to get their messages across. Even though political campaigns buy analytical reports about social media communities’ perception of candidates, which lead to – if necessary – campaign adjustments, there is no bi-directional communication between the campaign of the candidate and the social media communities. This is especially true for the elections in Germany.
The goal for this project is to analyze the perception of the political parties that participate in the German election in 2017 in the social media communities in near real time and visualize the results as a timeline during the course of the election year. As a first step, the sentiments of all German twitter messages (tweets) that address a political party will be analyzed and visualized using a developed web application. Furthermore, the user will be given the opportunity to interact with the most positive or negative tweets in order to enable a bi-directional interaction with the community. The user will also be given different options to calculate the sentiments with an explanation of these text-mining algorithms. In a second step, additional social media channels will be added to the stream.
The project web page and the web application are currently under development
October 2016 – October 2017
The project is funded by Saint Mary’s College of California