Early warning system for identifying critical situations in opinion formation on Web 2.0

Abstract: 

In Web 2.0, more and more people are joining together to form communities in which they exchange views about companies' products. In the process, they influence each other's opinions. By monitoring and analyzing the formation of opinions, companies can identify risks at an early stage.  

This paper presents an early warning system for identifying critical situations of opinion formation in Web 2.0. The system makes it possible to automatically analyze and assess the opinion formation process in communities. When a critical situation occurs, a warning is sent to the responsible marketing manager. The latter can avert negative consequences by intervening in time.  

The early warning system is designed as a neuro-fuzzy system. Based on the assessment of previous situations, it can learn linguistic rules that allow the assessment of future situations. Thus, it enables easy knowledge acquisition. Moreover, the rules are easily interpretable due to their linguistic form. Two use cases are used to show how the designed system can be used to identify critical situations in Web 2.0. 

This publication is only available in German.