The Semantic Knowledge Extractor Tool (SKET) is an unsupervised hybrid knowledge extraction system that combines a rule-based expert system with pre-trained machine learning models to extract cancer-related information from pathology reports.

The code is available in its GitHub repository.

Sket is currently used by the ExaMode partners to extract entities from medical reports and to classify them accordingly to the ExaMode categories. The results are used to train the weakly supervised learning framework.