BioNLP ontology extraction from a restricted language corpus with context-free grammars

BioNLP is an emerging area of NLP that brings new challenging objects for language processing and new valuable resources for bioinformatics and medicine. One notable task in BioNLP is creating de-novo ontologies. This is generally a tedious process; however, in some cases, it is possible to automate it to some extent. One such case is when a corpus of texts in a restricted subset of natural language is available. This paper presents a simple approach to automate ontology creation in such cases. The approach is aimed to simplify mapping of entities in natural texts to predefined ontologies wherever possible. The paper discusses which properties of the corpus enable the approach presented.