We describe a method for extracting data about how biomolecule pairs interact from texts. This method relies on empirically determined characteristics of sentences. The characteristics are efficient to compute, making this approach to extraction of biomolecular interactions scalable. |