Tuesday, September 10, 2019 - 12:15 to 13:15


Fine-grained Named Entity Recognition in Legal Documents

This paper describes an approach at Named Entity Recognition (NER) in German language documents from the legal domain. For this purpose, a dataset consisting of German court decisions was developed. The source texts were manually annotated with the following 19 semantic classes: person, judge, lawyer, country, city, street, landscape, organization, company, institution, court, brand, law, ordinance, European legal norm, regulation, contract, court decision, and legal literature. Overall, the dataset consists of approximately 67,000 sentences and contains around 54,000 annotated entities.

Open Government and the Semantic Web: a field report

Open Government — the right for citizens to access governmental information — has a long tradition in the Netherlands. Recent Dutch legislation requires government agencies to publish their data electronically, so that citizens can subscribe to information of their interest by e-mail. In order to facilitate this, we present a solution where we enrich government publication with knowledge graphs based on RDF/OWL. In this way, we can deal with the considerable variation in the ways that local governmental agencies enrich their data and publications.