-
Notifications
You must be signed in to change notification settings - Fork 72
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[FEATURE] Add z-score for the normalization processor #376
Comments
@samuel-oci Thanks for the feature request. Also welcome to take it and work on it. |
Sounds good, I'll take it, feel free to assign to me. |
Hi @heemin32 @vamshin @navneet1v I don't see any IT that tests normalization and was thinking to add those also as part of this PR. If I missed those or you already have those worked on somewhere else just let me know and we can consolidate. |
@martin-gaievski can you please provide the reference for the integration tests that has been added. I am able to find this: https://github.com/opensearch-project/neural-search/blob/main/src/test/java/org/opensearch/neuralsearch/query/HybridQueryIT.java#L32 |
@navneet1v this one test hybrid but for some reason when I run it in my IDE to debug I can't see the normalization processor being triggered. Could also be something wrong with my setup.. if it is then feel free to ignore. |
If you are using debugger it might not hit the code. I would recommend adding logs and check it. Debugger is tricky with IT tests, if you are just running |
@navneet1v I added the following lines
I do see logs of IT test and other core open search but don't see any logs showing from the normalization processor or NormalizationProcessorWorkflow. which is why I later run with debugger to double check. EDIT: Update: |
Is your feature request related to a problem?
Currently the normalization processor doesn't support z-score which is a popular technique and according in some instances produces superior results to min-max see blog
What solution would you like?
Allow to specify z-score as a normalization technique in the normalization processor
What alternatives have you considered?
Not much at the moment but suggestions are welcome.
Do you have any additional context?
see blog mentioned earlier
The text was updated successfully, but these errors were encountered: