Automatic rules extraction from medical texts

The majority of existing knowledge is encoded in unstructured texts and is not linked to formalized knowledge, like ontologies and rules. The potential solution to this problem is to acquire this knowledge through natural language processing (NLP) tools and text mining techniques. Prior work has focused on the automatic extraction of ontologies from texts, but the acquired knowledge is generally limited to simple hierarchies of terms. This paper presents a polyvalent framework for acquiring more complex relationships from texts and codes them in the form of rules. Our approach starts with existing domain knowledge represented as OWL ontology and SWRL "Semantic Web Rule Language" rules by applying NLP tools and text matching techniques to deduce different atoms as classes, properties etc. This is to capture the deductive knowledge in the form of new rules. We evaluate our approach thereafter by applying it on medical field more precisely Gynecology specialty, showing that this approach can generate automatically and accurately SWRL rules for the representation of more formal knowledge necessary for reasoning.