• de

Open Data Initiatives at HPI

This page gathers open-data-related activities at HPI, in particular in the area of linked open data and the semantic web. Contributing groups are

Research Projects

  • GovWILD
    GovWILD gathers and integrates government data from diverse government data sources, presents them in a Web interface and provides the data for download.
    Received the IBM Scalable Data Analytics Award
  • ProLOD
    Profiler for Linked Open Data (ProLOD) is a Web-based tool to perform data profiling tasks on datasets such as DBpedia. (NTII 2010 paper)
  • Yovisto
    Yovisto is a semantic video search engine specialized in academic content. Yovisto's search index is based on the combination of automated content based video analysis with user generated collaborative annotation (collaborative tagging, discussions, and comments) and Linked Open Data resources such as DBpedia. In difference to traditional video search engines, Yovisto enables pinpoint access within video data by providing fine-granular, time-dependent semantic metadata being published as Linked Open Data (SAMT 2010 paper / SemSearch 2010 paper).
  • Mediaglobe
    Cultural Memory provides an ever-increasing amount of information. Only a negligible small portion of this content is already accessable in the World Wide Web. Mediaglobe is a funded by the Federal Ministry of Economics and Technology within the context of the THESEUS research program 'New technologies for the internet of services' and aims to take media archives into the digital future providing availability and usability for the growing stock of audiovisual documents concerning German contemporary history. Mediaglobe offers state-of-the-art services for multimedia analysis and semantic multimedia search based on mappings to Linked Open Data resources on the web.
  • Stratosphere
    The DFG Research Unit (Forschergruppe) is developing a cloud-based data management system. One of the use cases is to query and cleanse linked open data.
  • iPopulator
    With iPopulator, we have introduced a system that automatically populates infoboxes of Wikipedia articles by extracting attribute values from the article's text.  In contrast to prior work, iPopulator detects and exploits the structure of attribute values to independently extract value parts. We have published a set of extracted facts from Wikipedia. (CIKM 2010 paper / technical report)


 ...  2010  2011  2012  2013  2014     Advanced Search

Profiling and Mining RDF Data with ProLOD++

Ziawasch Abedjan, Toni Gruetze, Anja Jentzsch, Felix Naumann
In Proceedings of the IEEE International Conference on Data Engineering (ICDE), Demo, Chicago, IL, 2014

Assigning Global Relevance Scores to DBpedia Facts

Philipp Langer, Patrick Schulze, Stefan George, Matthias Kohnen, Tobias Metzke, Ziawasch Abedjan, Gjergji Kasneci
In International Workshop on Data Engineering meets the Semantic Web (DESWeb) , Chicago, IL, 2014

Amending RDF Entities with New Facts

Ziawasch Abedjan, Felix Naumann
In Know@LOD Workshop in conjunction with ESWC, Creete, Greece, 2014 Selected for Best Workshop Paper Award.
 ...  2010  2011  2012  2013  2014 


Master's Theses

  • Learning to Extract Structured Information from Wikipedia Articles to Populate Infoboxes (Dustin Lange, 2009) (see also CIKM 2010 paper and tech report)
  • Wikipedia cross-lingual Concept Identification and Infobox Alignment (Daniel Rinser, 2010)
  • Strategies for structure-based rewriting of SPARQL queries for data prefetching (Armin Zamani, 2013)