Skip navigation links


See: Description

Package Description

This package contains the older version of Stanford NER tagger that uses a Conditional Markov Model (a.k.a., Maximum Entropy Markov Model or MEMM) designed for Named Entity Recognition, and various suppport code. For newer version of the Stanford NER tagger that is based on Conditional Random Fields (CRF) models, please see The original version was written for the CoNLL 2003 Shared Task competition. The design was considerably generalized, and many more features added for the BioCreative 2003 evaluation. For basic NER tagging, you should use CMMClassifier. (Earlier versions still exist as NGramStringClassifier and BioNGramStringClassifier.) The rest of the code in this directory is various remnants of our BioCreative system and various support utilities for NER or CMMClassifier.

Skip navigation links

Stanford NLP Group