Japanese named entity recognition (NER) models/data?

Hi

The package here: http://l2r.cs.uiuc.edu/~cogcomp/asoftware.php?skey=FLBJNE is a publicly available tool that is one of the best-performing on the English CoNLL 03 shared task.

To use it, you'll have to run Percy Liang's Brown clustering algorithm on a large collection of unlabeled Japanese text, and to build some gazetteers from the Japanese version on Wikipedia. I hope this helps

A number of approaches exist for this. You can look at any of the SVM based approaches (for example, this paper that uses structure information in various forms), or other supervised approaches such as ones that use decision lists, or decision trees.