26th Annual International Symposium on Computer and Information Science, London, Kanada, 26 - 28 Eylül 2011, ss.527-532
Turkish is an highly agglutinative language and this poses problems in information retrieval from Turkish free-texts. In this paper one rule based and one data driven alternate methods for information retrieval from Turkish radiology reports are presented. Both methods do not utilize any medical lexicon or ontology, although inherent medical information within the training reports exist. Performance is measured and evaluated as a retrieval problem.