TeMek – Text Mining of Message Texts for Standardized Classification
With 200,000 deaths annually, cancer is Germany’s second most common cause of death. A central component in the fight against
cancer is the collection and provision of high-quality data on the diseases that occur in Germany by the cancer registries of the German federal states. The goal of the TeMeK project (text mining of message texts for standardized classification) is to facilitate the processing of free-text cancer findings from the clinical tumor centers within the state cancer registries by text mining, by automatically extracting the required structured cancer information from the free texts. The interdisciplinary project team consisting of cancer registries, clinical tumor centers and development partners (the Fraunhofer Institute FKIE and Averbis) is trying to adapt the already existing text mining tools to support tumor documenters at cancer registries and to adapt them to new legal requirements for tumor documentation. Averbis can already draw on more than 7 years of experience in cooperation with the state cancer registries for text mining-driven structuring of findings.