Changes between Version 10 and Version 11 of WikiStart


Ignore:
Timestamp:
03/18/09 09:15:32 (15 years ago)
Author:
horak
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • WikiStart

    v10 v11  
    1 ''' This site is under heavy construction! ''' 
    2  
    3 = Intelligent Document Information Extraction = 
     1= iDocument: Intelligent Document Information Extraction = 
    42[[Image(WikiStart:logo.png, width=100px, right)]] 
    53 
    6 iDocument is an [http://en.wikipedia.org/wiki/Information_extraction| information extraction] system. It uses existing background knowledge such as business databases, concept maps, or other information models for improving extraction results. 
     4iDocument is a  generic  ontology-based information  extraction  (OBIE)  system  that  uses  ontological  background  knowledge  in  
     5terms  of  existing  vocabularies  and  instance  knowledge.  iDocument  uses  existing knowledge  from personal or business domains (e.g. relational databases, concept maps, taxonomies,  etc.).  Following  Semantic  Web,  iDocument  exchanges  and  extracts  
     6knowledge based on  the W3C  standard RDF. Existing knowledge  is used as  input  in a serial  IE  pipeline  of  extraction  tasks  for  extracting  possible  answers  concerning  user specified  ad  hoc  queries  on  a  given  text  collection.  
    77 
    8 It extracts entities (eg.,email adresses, names, URLs), instances, and semantic relations between instances with respect to used background knowledge. iDocument uses ontological knowledge representation techniques for interpreting background knowledge.  
     8= Unique Feature = 
    99 
    10 [[Image(WikiStart:scenario.png, center)]] 
     10 * Domain ontologies are exchangeable as long as they are written in RDFS. 
     11 * The MOBIE mapping vocabulary allows to define relevant classes, attributes and relations for extraction purpose . 
     12 * Existing instance knowledge is reused for information extraction purpose 
     13 * Extracted results are formalized in the same RDF scheme as the input domain ontology. 
     14 * SPARQL queries are used for defining extraction templates. 
     15 * All intermediate and final extraction results are weighted hypothesis according to Dempster –Shafer’s belief function.  
     16 
     17 
     18= Table of Contents = 
     19 
     20 * [http://idocument.opendfki.de/ System Summary] 
     21 * [http://idocument.opendfki.de/wiki/Evaluation/Corpus/OlympicGames2004 Olympic Corpus and Annotation Scheme (OCAS)] 
     22 * [http://idocument.opendfki.de/ Publications] 
     23 * [http://idocument.opendfki.de/ Reference Projects] 
    1124 
    1225For further information please contact [mailto:benjamin.adrian@dfki.de].