Detecting Arguments and Their Positions in Experimental Communication Data

This paper applies argument mining techniques to political discussions obtained from an online chat experiment prior to a ballot. We investigate the task of automatically detecting chat messages that give justification for an underlying claim. We use bag-of-words features as well as state of the art embedding models to train different classifiers on the given task. Moreover, we automatically detect each argument's position and compare our method to a method based on word scores proposed in the political science literature. Our results show that argument mining techniques can be successfully applied to experimental communication data and that it can replace labor-intensive manual annotations by human coders