ConText, a widely used, open-source clinical natural language processing algorithm, tries to confirm that information extracted from a clinical note applies to the appropriate patient and visit and is not negated. The process, however, is time consuming as it must follow over 600 rules.

FastConText is a more efficient, scalable implantation of ConText suitable for large- scale clinical natural language processing. The algorithm determines contextual features of information from clinical notes by identifying negation, temporality, and experience using generalized rule processing. The new algorithm adds additional rules that improve both the speed and accuracy of natural language processing.