作者: Ivelina Nikolova , Dimitar Tcharaktchiev , Svetla Boytcheva , Zhivko Angelov , Galia Angelova
DOI: 10.1007/978-3-319-10554-3_9
关键词: Big data 、 Public health policy 、 Text messaging 、 Computer science 、 Bulgarian 、 Medical emergency 、 National health insurance 、 Business intelligence 、 Health care 、 Drug treatment
摘要: This paper presents a research project integrating language technologies and business intelligence tool that help to discover new knowledge in very large repository of patient records Bulgarian language. The ultimate objective is accelerate the construction Register diabetic patients Bulgaria. All information needed for available outpatient records, collected by National Health Insurance Fund. We extract automatically from records’ free text essential entities related drug treatment such as names, dosages, modes admission, frequency duration with precision 95.2%; we classify according hypothesis “having diabetes” 91.5% deliver these findings decision makers order improve public health policy management healthcare system. experiments are run on about 436,000 patients.