I am a Research Assistant in DERI at the UNLP group led by Paul Buitelaar. My research interests are the use of domain vocabularies in Information Extraction and their lexical enrichment across languages. Finding Semantic Web compliant standards to represent these lexicons, which go beyond the capabilities of SKOS is another challenge.
I studied Computer Science at the University of Ulm, Germany where I did my Diploma Thesis in the Semantic Web Group (under supervision of Dr. Thorsten Liebig and Prof. Dr. Friedrich von Henke). In my thesis (“Enhancing the Relation Extraction Process from Natural Language with Linguistic Enriched Domain Ontologies”) I developed an IE (Information Extraction) system to extract relationships from Natural Language (NL) German News Texts. The prototype was based on a modular architecture using (i) a domain knowledge base with a linked (ii) lexicon. Another aspect of my thesis was how to measure semantic relatedness of different relationships in the domain ontology (e.g. isCEO vs. isHeadOfCompany).
I started my work at DERI in September 2009 working on the LexInfo project. LexInfo is a implementation of the lemon (lexicon model for ontologies) model to generate domain specific lexica from ontology models to enrich them with semantic (graph patterns), terminological (variants/multilingual) and linguistic (morphology/syntax) information.
In 2010 I joined the MONNET (Multilingual Ontologies for Networked Knowledge) project as a PhD student. The project aims to develop standards to interchange linguistic information on the Semantic Web for Information Extraction and Translation applications. In particular I worked on Cross-Lingual Ontology-Based Information Extraction (CLOBIE) in the financial domain.
In 2012 I joined UNLP as Research Assistant where I am working on Multilingual Semantic Applications within the MONNET project.