Cool Semantic Web projects at EIS and Knowledge Graphs

Date: Feb 10, 2017

Speaker: Michael Galkin, University of Bonn

Abstract

Michael will present a few projects developed in the Enterprise Information Systems (EIS) department at the University of Bonn, and give short demos of the following systems:

1. OpenResearch http://openresearch.org/Main_Page - collaborative scientific information platform
2. VoCol http://butterbur06.iai.uni-bonn.de/ - collaborative ontology development
3. SlideWiki http://slidewiki.org/ - collaborative online presentation tool
4. ODINE - https://opendataincubator.eu/ - open data incubator for Europe

Michael will also talk about a large project they are working on:

• WDAqua - http://wdaqua.eu/ - Open Question Answering platform
• If there is time, Michael will also discuss his PhD topic - Enterprise Knowledge Graphs on top of Semantic Data Lakes.

Data Lakes have been proposed to provide scalable and flexible data discovery, analysis, reporting and to reduce the initial integration efforts before the data sources can be used.

Although Data Lakes reduce the costs of identifying, storing, cleansing, and integrating data significantly and promote flexibility in data analysis, they introduce complexity in exploring, querying, and analyzing the available raw data in a unified manner. In this work, we introduce an ontological architecture for Data Lakes (Semantic Data Lake, SDL) tailored for large-scale heterogeneous data.

SDL leverages mappings to a knowledge graph to provide a robust and universal query mechanism with SPARQL as a common query language.

SDL outperforms existing SPARQL query federation approaches on both synthetic and real datasets on the scale of hundreds of millions of triples.