• deutsch | english

Context changes in Social Media Contributions


Today, social media are ubiquitously available and change traditional patterns of communication. This influences journalism as well, where social media are used increasingly, to keep pace with the dynamicity and volatility of this medium.
However, the increase in speed and the format of social media messages also poses a number of new problems. One issue is information that undergoes context changes or is put out of its original context. This might happen unintentionally through shorting or deliberately, as the ongoing discussion on `Fake News’ and attempts to manipulate public opinion indicate.
This joint project of the IWM and the University of Stuttgart aims at the development of novel approaches to detect and analyze changes of context interactively. For this, the project will develop detection methods based on contemporary natural language processing for detecting (candidate) context changes. Interactive visual approaches will complement these methods, since nuances that are difficult to recognize automatically can cause big semantic differences. The project has different goals. On the one hand it will enable users to adapt algorithmic detection methods to new forms of context changes and on the other, it will help users to analyze and track the temporal evolution of context changes. Finally, the project will evaluate whether the developed approaches outperform traditional ways of detecting changes in context and how well different user groups can be supported by them.
The researchers of the IWM and University of Stuttgart complement each other with respect to research expertise for this project. The Institute for Visualization and Interactive Systems of the University of Stuttgart will develop the detection methods and visual approaches, while the IWM will conduct studies on detecting context changes and evaluate the proposed visualization approaches.