A student project inspired by real research.
From Classroom to Conference
Hands-On Research Experience in the Classroom
During the winter semester of 2025/26, fifth-semester students in the General and Digital Forensics program at Mittweida University got hands-on experience with forensic text mining in a software project. In small teams, they worked on detecting harmful content on social media, using tweets as their data. Instead of focusing only on hate speech, the students explored less common research topics, like spotting calls for violent action and figuring out if hateful tweets targeted specific groups.
The project was designed to closely resemble the workflow of scientific competitions known as shared tasks. Students not only developed and evaluated classification systems but also documented their approaches and results in system papers and underwent a scientific review process.
Final Conference and Poster Session
The project ended with an internal academic-style conference on February 12, 2026, where students presented posters and discussed their work over coffee and cake.
A highlight was the Best Poster Award, chosen by students and lecturers, which recognised an eight-member team for their creative and appealing poster presentation on their ensemble learning approach to target recognition. Congratulations to the winners! 
Looking Ahead
Students who discovered an interest in research through this project can look forward to the next opportunity: the second edition of the Harmful Content Detection Shared Task, which will be organized as part of GermEval 2026 at KONVENS 2026. Three of the tasks the students explored will be part of this nationwide competition. We look forward to welcoming our students and many new participants back for these exciting upcoming challenges!
Prof. Dr. rer. nat. Michael Spranger NEWS
2026 conference