CANE:Corpus-Adaptive Named Entity Recognition (PhD-Thesis 09/2002-12/2006)
An approach based solely on a sophisticated interaction of labeled and unlabeled data in order to facilitate the fast development of robust systems for Named Entity Recognition in challenging domains.
PROMETHEUS (01/2004-03/2005)
Prometheus is an intelligent annotation environment designed for interactive Information Extraction robust to OCR-scanned texts.
Named Entity Recognition for E-Science - subproject of WIKINGER (10/2005-9/2008) together with Andreas Wagner
Within the BMBF-funded project WIKINGER we are currently working on robust and easy to develop services for Named Entity Recognition. The pilot domain of our project is Contemporary History of Catholicism. In cooperation with the domain experts of the KFZG appropriate categories are explored and defined. For our learning based recognition algorithm, we are working on techniques to reduce the effort of manually labelling data. Both, the discussion and definition of the Named Entity categories and our research on reducing the labelling effort are tightly coupled with WALU.
WALU Wikinger Annotier- und Lernumgebung - subproject of WIKINGER (10/2005-9/2008)
together with Andreas Wagner
WALU is an environment currently under development for the annotation and learning of Named Entities. WALU is designed to facilitate the creation of labelled training data by supporting the human annotator, to setup, maintain and coordinate recognition modules and to support the collaborative process of finding, discussing and defining Named Entity categories useful for a given domain.
Publications
Lars Bröcker, Marc Rössler, Andreas Wagner. "Knowledge Capturing Tools for Domain Experts". In: SAAKM 2007 - Semantic Authoring, Annotation and Knowledge Markup Workshop. Co-located with the 4th International Conference on Knowledge Capture (K-Cap 2007), Whistler, British Columbia, Canada, October 28-31, 2007. [to appear]
M. Rössler, A. Wagner, F. Jungermann, W. Hoeppner. "Applying WALU to Annotate Named Entities in Italian". Shared task contribution. In: Intelligenza Artificiale - Periodico trimestrale dell' Associazione Italiana per l' Intelligenza Artificiale (http://ia.di.uniba.it/), Proceedings of EVALITA 2007 - Evaluation of NLP Tools for Italian, workshop organized in conjunction with AI*IA 2007 (http://aiia.info.uniroma2.it/), Roma, Italy, September 10-13, 2007.
Lars Bröcker, Stefan Paal, Andreas Burtscheidt, Bernhard Frings, Marc
Rössler, Andreas Wagner, Wolfgang Hoeppner.
"WIKINGER - Wiki Next Generation Enhanced Repositories". German e-Science Conference 2007. Baden-Baden, Germany, 2007. [pdf]
Andreas Wagner und Marc Rössler. "WALU – Eine Annotations- und
Lern-Umgebung für semantisches Tagging". GLDV-Frühjahrstagung, Tübingen, 2007. [pdf]
Marc Rössler. "Corpus-based Learning of Lexical Resources for German Named Entity Recognition". In: Proceedings of LREC 2004. Lisboa, Portugal, 2004. [pdf]
Marc Rössler. "Adapting an NER-System for German to the Biomedical Domain". Shared task contribution. International Joint Workshop on Natural Language Processing in Biomedicine and its Applications - Geneva, Switzerland, 2004. [pdf]
Marc Rössler. "Using Markov models for named entity recognition in German newspapers". In: Proceedings of the ESSLLI'02 Workshop on Machine Learning Approaches in Computational Linguistics. Trento, 2002. [pdf]
Marc Rössler. "Zur Erschliessung historischer Ressourcen mittels computerbasierter Medien". Magisterarbeit. Gerhard Mercator Universität Duisburg, 2000.
Teaching
Praxisprojekt Kommedia (SS 06)
Logics und Data Structures, Lecture and Exercise (SS 05)
Machine Learning for Natural Language Processing (WS 04/05)
Grammar Formalisms and Parsing (WS 04/05)
Algorithms and Formal Languages, Lecture and Exercise on Finite State Automata (SS 04)
Information Extraction (SS 04)
Introduction to Perl, Lecture and Exercise (WS 03/04)
Logics und Data Structures, Lecture and Exercise (SS 03)
Named Entity Recognition, Guest Lecture at the University Joensuu, Finland (March 2003)
Introduction to Perl (WS 02/03)
Grammar Formalisms and Parsing (WS 02/03)
Information Extraction and Text Mining (SS 02)
Algorithms and Formal Languages, Exercise (SS 02)
Searchengine Spotting (WS 01/02)
Information Science for Computational Linguists, Lecture and Exercise (WS 01/02)
Logics und Data Structures, Lecture and Exercise (SS 01)
Computational Linguistics, Markup Languages, and the Internet (WS 00/01)
