Word tagging evaluation drives developments in downstream technologies such as named entity detection, which, in turn, drives high-end applications such as coreference. You will notice that much of the evaluation resembles the evaluation from our classifiers except that each tag is evaluated like its own classifier category.
This recipe should serve to get you started on evaluation, but be aware that there is a very good tutorial on tagging evaluation on our website at http://alias-i.com/lingpipe/demos/tutorial/posTags/read-me.html; this recipe goes into greater detail on how to best understand tagger performance.
This recipe is short and easy to use, so you have no excuses to not evaluate your tagger.