CODEWITHSUNDEEP
javamachine-learningnlpstanford-nlpnamed-entity-recognition
BlueGirl
BlueGirl

Reputation: 501

Number of Tags of NER

As far as I know, Stanford NER has 3,4 and 7 class models (or tags). I need a standard and a Java implementation of a Named Entity Recognizer which has more than 7 tags (for example 13 tags). I don’t want to develop it by myself and I need it to be accurate. Is there any NER which satisfies these conditions?

Thanks

Upvotes: 1

Views: 1493

Answers (2)

Rahul K P
Rahul K P

Reputation: 16081

Python nltk package provide Named Entity Recognition(NER), it have an in build classifier. It have more than 7 classes. Which are

ORGANIZATION - Georgia-Pacific Corp., WHO
PERSON - Eddy Bonte, President Obama
LOCATION - Murray River, Mount Everest
DATE - June, 2008-06-29
TIME - two fifty a m, 1:30 p.m.
MONEY - 175 million Canadian Dollars, GBP 10.40
PERCENT - twenty pct, 18.75 %
FACILITY - Washington Monument, Stonehenge
GPE - South East Asia, Midlothian

You are using Java,so there is an equivalent package in java which is OpenNLP . You can check the classifier in OpenNLP.

Here is the SO link which proves OpenNLP is equivalent to nltk

Upvotes: 1

Gabor Angeli
Gabor Angeli

Reputation: 5749

The Stanford NER system can be retrained for any number of tags. The 3, 4, and 7 class models are just the pre-trained models. If you'd like to train it on more classes, you'd have to collect training data with those classes in it, and retrain the model.

Upvotes: 1

Related Questions