# Mapping Toxic Comments Across Demographics: A Dataset from German Public Broadcasting

## TL;DR

This paper introduces a large-scale German dataset of online comments annotated for toxicity and demographic information, enabling analysis of age-related differences in toxic speech for improved moderation.

## Contribution

It presents the first German toxicity dataset with demographic annotations, combining human and language model annotations, to study age-based toxic communication patterns.

## Key findings

- Younger users use more expressive language in toxic comments.
- Older users more frequently spread disinformation and devalue content.
- The dataset enables demographic-specific toxic speech analysis.

## Abstract

A lack of demographic context in existing toxic speech datasets limits our understanding of how different age groups communicate online. In collaboration with funk, a German public service content network, this research introduces the first large-scale German dataset annotated for toxicity and enriched with platform-provided age estimates. The dataset includes 3,024 human-annotated and 30,024 LLM-annotated anonymized comments from Instagram, TikTok, and YouTube. To ensure relevance, comments were consolidated using predefined toxic keywords, resulting in 16.7\% labeled as problematic. The annotation pipeline combined human expertise with state-of-the-art language models, identifying key categories such as insults, disinformation, and criticism of broadcasting fees. The dataset reveals age-based differences in toxic speech patterns, with younger users favoring expressive language and older users more often engaging in disinformation and devaluation. This resource provides new opportunities for studying linguistic variation across demographics and supports the development of more equitable and age-aware content moderation systems.