Fine-Grained Named Entity Recognizer

What is FG_NER?

al+ FG_NER is a Fine-Grained Named Entity Recognizer by Alt Inc.
It recognizes 200 fine-grained named entity categories, which is a huge extension of the traditional 6~8 categories

Existing Named Entity Recognizers
6 ~ 8 Categories

6~8 Categories

al+ FG_NER
200 Categories With Hierarchical Structure

200 Categories With Hierarchical Structure

Examples of Use

Existing Named Entity Extractors

Existing Named Entity Extractors



Performance (F-score) Comparison

Terminal - Cloud Type

Test Data Mapping And Corresponding Evidence:see this link

Test/Comparison data : Stanford NER based

Test/Comparison data : Google NLP API based

For Trial of FG_NER

Demo version :

Terminal - Cloud Type

We have prepared an environment such that you can try out the different NERs
to see their differences with al+ FG_NER instantaneously

FG_NER Model Builder (A System which narrows down your required categories)

Terminal - Cloud Type

With this system, you can generate new models which narrows down your required categories in accordance with its application purpose. The FG NER’s average accuracy of 80% is further enhanced with 200 categories

Need to register an al+ account.
For trial version, please contact

Contact us

For business, services or joint development matters, please contact us at