-1

I wanted to develop a model using Watson Knowledge studio, which has both rule based and machine learning models(some entities must be rules based, some must be machine learning pre-annotator), due to limitations from Watson i cannot build a single model, if i deploy both the models in same NLU instance, is it possible for me to use both model id's in same API call?

running 2 different API's calls for same set of documents is not economical.

Please suggest if there is any other solution supporting my issue.

thank you

desertnaut
  • 57,590
  • 26
  • 140
  • 166
Priyanka Athawale
  • 1
  • 1

2 Answers2

1

It is not supported to use both model in same NLU API call. You can send your request of new feature on https://ibm-watson.ideas.aha.io/?project=WKS.

Takasaki Fukai
  • 11
  • 1
0

Create a custom model for your business need. That way you can use ML and RegEx(Rule based) models together , in an overlapping pattern. Imagine this as two separate models are running parallel to each other and annotating same set of documents.

If you want you can also compare your custom model to NLU OOB model to evaluate performances.

Dyuti Abha
  • 1