Faculty of Informatics and Statistics, Department of Information and Knowledge Engineering (DIKE)

Date and time: December 8 2011 (10:30 – 12:00). Non–standard date or time!

Room: 403 NB


Complementing the Semantic Web with Machine Learning


  • Uwe Stoll, University of Bundeswehr, Munich, Germany

Problem definition: Given the growing amount of e-commerce websites annotated with structured data, how can we use this data to design a Machine Learning system that is capabable of extracting structured data from non-annotated websites with high precision & recall and with minimal computational effort.
I will go into details of extraction targets, data signals usable for the task, and what I achieved so far.

Powered by Resource Description Framework (RDF)