Jugendschutz im Internet

Development of Innovative Technologies for the Identification of Internet Content that is Harmful to Adolescents 


The company ComNet GmbH, in collaboration with the HHU Düsseldorf (Institute of Language and Information, Department of Computational Linguistics), TU Dresden (Institute of Acoustics and Voice Communication) and the company Cognesys GmbH, will develop software for automatic filtering of text and video content on the Internet. The aim is to provide an application that goes beyond building blacklists or identification of single keywords and understands and evaluates the complete content of a webpage. 

The main module of the application will be based on Cognitive Ergonomic Solution (CES) of Cognesys GmbH, a universal speech and text interface for automatic acquisition und further processing of the meaning of spoken and written information.

The performance of CES can be significantly enhanced if the input is annotated with lemmata and relevant morphosyntactic information. Our contribution to the project will be to develop broad-coverage lemmatizers for German and English.

Each parser will consist of the following language-specific modules: tokenizer, sentence splitter, normalizer, morphological analyzer (lexicon) and simple phrase tagger. The last module is needed for the disambiguation of homonymous forms belonging to different part-of-speech classes (e.g., books (noun plural) and books (verb, 3rd person singular, present tense)) and for linking split verb forms in German and phrasal verbs in English in order to provide the correct lemma (e.g., aufhören for Hör auf!).