We have collected a large set of tweets observing the German federal election in 2017. Tweets originate from candidates to the German parliament, from different media accounts and from people who tweeted using a prespecified set of hashtags. Some of the account holders are described by metadata specifying, e.g. their party or their candidate status. From a political science perspective it would be interesting to find out whose behavior deviates from those of others. A subtopic in machine learning is subgroup discovery. It tries to determine a subgroub of items (here: twitter accounts) whose characteristics is different than those of others. In this thesis one would have to research and explore whether or how subgroup discovery has been applied to social media in the past, what the findings were and whether such findings can be carried over to our above setting. The interestingness of subgroups would have to be discussed within our project E-Democracy with political scientists.
- Herrera, F., Carmona, C.J., González, P. et al. Knowl Inf Syst (2011) 29: 495. https://doi.org/10.1007/s10115-010-0356-2 http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.646.4342&rep=rep1&type=pdf
- Lavrač, N., Kavšek, B., Flach, P., & Todorovski, L. (2004). Subgroup discovery with CN2-SD. Journal of Machine Learning Research, 5(Feb), 153-188. http://www.jmlr.org/papers/volume5/lavrac04a/lavrac04a.pdf