Social media monitoring has become an important means for business analytics and trend detection, for instance, analyzing the sentiment towards a certain product or decision. While a lot of work has been dedicated to analyze sentiment for English texts, much less effort has been put into providing accurate sentiment classification for the German language. In this paper, we analyze three established classifiers for the German language with respect to Facebook posts. We then present our own hierarchical approach to classify sentiment and evaluate it using a data set of ~640 Facebook posts from corporate as well as governmental Facebook pages. We compare our approach to three sentiment classifiers for German, i.e. AlchemyAPI, Semantria and SentiStrength. With an accuracy of 70 %, our approach performs better than the other classifiers. In an application scenario, we demonstrate our classifierl’s ability to monitor changes in sentiment with respect to the refugee crisis.
|Name||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Conference||21st International Conference on Applications of Natural Language to Information Systems|
|Period||22/06/16 → 24/06/16|
- Information, Communication & Computing