There are several good taggers around. I even asked a question creating own tagger,
I have got another requirement now. In Python I was using topia and it seemed a great choice for job (fast and concise). But there is no such alternative in Java,I could find.
Now, I have three questions related to this :
1)Is there any term extractor/pos tagger in Java which is based on FSM?
2) Is FSM tagger "CAN BE" more efficient (I know it is way faster, but accuracy) than corpus based taggers?
3) How Do I start building One in Java? Any basic guide creating machine extracting pos tags from sentence :- "Einstein was a great scientist." ? Just a start ?
来源:https://stackoverflow.com/questions/16371483/any-fsm-fsa-based-tagger