Extracting process graphs from medical text data: An approach towards a systematic framework to extract and mine medical sequential processes descriptions from large text sources

Abstract

In this paper a natural language processing workflow to extract sequential activities from large collections of medical text documents is developed. A graph-based data structure is introduced to merge extracted sequences which contain similar activities in order to build a global graph on procedures which are described in documents on similar topics or tasks. The method describes an information extraction process which will, in the future, enrich or create knowledge bases for process models or activity sequences for the medical domain.