Overview
In turbulent times, capturing current sentiments and opinions is more relevant than ever for companies to identify trends, launch new campaigns, make investment decisions or develop new products. In social media, but also on online media platforms, every day many posts are written and commented on, which in turn generates a wealth of text data. This language data can be analysed according to content, valence, and emotional intensity. Valence measures sentiment, i.e. how positive or negative a text is. Emotional intensity measures how strong emotions are in a text. Valence and emotional intensity are central to the early detection and prediction of scandals. Certain companies polarise and repeatedly find themselves in scandals, which damages their own reputation but also has a negative impact on partner companies. Other firms, on the other hand, are more neutral and generally well-liked. A company's own reputation is one of its most important resources. It is therefore important to protect it. The aim of this project is to test the idea of an automated real-time sentiment barometer that uses existing language data from various online sources to examine the reputation of companies.
