ONTOSOPHIA - Ontology-driven Information Extraction Supported by (Semi-) Automatic Corrective Feedback

  • Christina Sachsenhofer-Feilmayr (Speaker)

Activity: Talk or presentation › Poster presentation › unknown

Description

Existing (Web-) information extraction (IE) systems are domain dependent or limited to documents with an underlying common structure. The deployment of real applications shows, that permanent maintenance is essential in order to improve extraction quality and to cope with sustainability issues. ONTOSOPHIA is the development of a novel domain independent approach of IE, enabling domain experts in creating and maintaining a domain dependent IE system, that is able to extract complex templates from (Web) documents with marginal involvement of knowlegde engineers. The upcoming approach of ontology-driven information extraction (ODIE) will be extended by integrating a corrective feedback component that (semi-) automatically allows for the maintenance of the extraction domain ontology.
Period24 Oct 2009
Event titleISWC Doctoral Consortium
Event typeConference
LocationUnited StatesShow on map

Fields of science

  • 502007 E-commerce
  • 102006 Computer supported cooperative work (CSCW)
  • 102028 Knowledge engineering
  • 102016 IT security
  • 102019 Machine learning
  • 509018 Knowledge management
  • 506002 E-government
  • 505002 Data protection
  • 102022 Software development
  • 102033 Data mining
  • 102010 Database systems
  • 102035 Data science
  • 102001 Artificial intelligence
  • 102015 Information systems
  • 202007 Computer integrated manufacturing (CIM)
  • 102014 Information design
  • 102025 Distributed systems