AI-supported collective-social moderation of online discourses (KOSMO)
In the last decade, digital platforms have become important instruments for the joint development of ideas and places for the exchange of political opinions. However, considerable human and financial resources are required to moderate and evaluate the many contributions.
Using state-of-the-art machine learning technologies and gamification approaches, the KOSMO project has developed a prototype for a system that proactively supports moderators in carrying out their tasks and thus enables administrations, political institutions, non-governmental organizations (NGOs), foundations and online media in particular to conduct cost-effective but high-quality online discussions and online participation processes. This should make the discussions more attractive for participants and providers and easier to evaluate. In this way, KOSMO enables the implementation of online discussions and digital participation processes on a broad scale.
From a scientific perspective, KOSMO pursued three goals:
- The optimization of algorithms for the automated detection of low-quality and high-quality
- discussion contributions for the German language.
- The integration of the algorithms in a prototype of an assistance system that proactively supports moderators in their tasks.
Provide in-depth answers to the research questions of how online discussions can be improved by assistance systems for semi-automated moderation and synthesis, how these systems are perceived and evaluated by providers and participants of online discussions and how participation processes can contribute to strengthening democracy through the use of the systems.
The project was carried out in cooperation with the DIID, Liquid Democracy e.V. (LIQD), the Institut für Partizipatives Gestalten / Hörster & Rohr GbR (IPG) and the Deliberative Discussions on the Social Web (DEDIS) research group at Heinrich Heine University Düsseldorf. Further information can be found here.
The Düsseldorf sub-project was funded by the Federal Ministry of Education and Research (BMBF) with around €370,000 for the period from April 2020 to March 2022.
You can find more information on the official project website or in the final report (see below).
Contact
Prof. Dr. Marc Ziegele (Speaker)