GROBID is a machine learning library for extracting, parsing and re-structuring raw documents such as PDF into structured XML/TEI encoded documents with a particular focus on technical and scientific publications. First developments started in 2008 as a hobby. In 2011 the tool has been made available in open source. Work on GROBID has been steady as a side project since the beginning and is expected to continue as such. - View it on GitHub