Automatic Knowledge Extraction Supported by Semantic Enrichment in Medical Records

Abstract

The volume of digital information is growing considerably in the last two decades and there is currently a huge concern about obtaining this content quickly and effectively. In the health sector it is not different; to retrieve medical records that obtain relevant information about treatments and progresses of clinical conditions may speed up new patients' diagnosis. In this work it is described a framework proposed for automatically indexing information based on semantics and on text mining techniques. This task should work in parallel with the insertion of data into electronic records. The original contributions come down to search engine in texts organized so as to potentiate the amount of results obtained, as evidenced by the experiments carried out. The stored information is automatically fragmented into words, which have a semantic dictionary based on a framework that enables the information retrieval through semantics.