Future Work (1/2)
		There’s a lot of room for improvement on this work. E.g., standard things like: • Using ontological information • Adding more training data • More careful coding of –  Textual patterns: / \digits <NN> / for cohort features –  Position: finer grained line position and general section determination –  Orthography: case, punctuation and number information –  Named Entities: and co-reference chains