DocumentCode :
3174793
Title :
Similarity estimation among OWL descriptions of computational cardiology problems in a knowledge base
Author :
Bratsas, Ch ; Hatzizisis, I.S. ; Bamidis, P. ; Quaresma, P. ; Maglaveras, N.
Author_Institution :
Aristotle Univ., Thessaloniki
fYear :
2005
fDate :
25-28 Sept. 2005
Firstpage :
243
Lastpage :
246
Abstract :
The purpose of this study is to evaluate ontologies in order to achieve an efficient search of medical computational problems in cardiology (MCP). These ontologies provide the ability to standardize the description process of cardiology problems and algorithms by utilizing Ontology Web Language -Description Logic (OWL-DL). The final aim is to build a knowledge base (KB) from the OWL descriptions of the MCPs. The identification of a new cardiology problem or related to this problem included in KB is proposed to be conducted by cosine-based similarity. A novel algorithm that constructs the MCPs vectors from the ontology individuals is presented. These vectors are defined by a matrix based on problems, Concept Unique Identifier (CUI) Keywords and CUI Categories. The CUIs are automatically received by the Unified Medical Language System Knowledge Server (UMLS-KS) and stored into ontologies. Finally, an example where twelve cardiology problems vectors are constructed from the corresponding individuals of the ontology classes is presented and analyzed
Keywords :
cardiology; knowledge based systems; knowledge representation languages; medical computing; ontologies (artificial intelligence); computational cardiology problem; concept unique identifier; knowledge based ontologies; medical computational problems; ontology Web language-description logic; unified medical language system knowledge server; Biomedical imaging; Cardiology; Knowledge engineering; Logic; OWL; Ontologies; Resource description framework; Scattering; Software libraries; Unified modeling language;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computers in Cardiology, 2005
Conference_Location :
Lyon
Print_ISBN :
0-7803-9337-6
Type :
conf
DOI :
10.1109/CIC.2005.1588082
Filename :
1588082
Link To Document :
بازگشت