Modular Ontologies
Rule Extraction from Support Vector Machines
Rule Extraction from Support Vector Machines
| Hi-index | 0.00 |
The development of clinical ontologies using common clinical data is a very important issue to record healthcare patient history, to use medical guidelines and to services accountability. The usage of terminologies already developed and available like SNOMED is a benefit. However many doctors argue that they prefer to continue using natural language and unstructured text to record patient data. Their point of view is that natural language is much more complete and flexible than standardized terminologies. This study intends to prove that it is possible to recognize patterns from natural language and identify the clinical procedures as they would be written with a normalized language. Another delivery of this study would be a precisely accountability of healthcare services.