Skip to content

Search Request Processor: pipeline for transformation of queries and results inline with a search request.

License

Notifications You must be signed in to change notification settings

ivan-tse/search-processor

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Build and Test Search Query & Request Transformers codecov PRs welcome!

Search Rerankers: AWS Kendra & AWS Personalize

Welcome!

This repository hosts the code for two self-install re-rankers that integrate into Search Pipelines. User documentation for the Personalize Reranker is here. For Kendra, it is here.

Search Processors: Where Do They Go?

The current guideline for developing processors is that if you are developing a processor that would introduce new dependencies in OpenSearch Core (e.g. new libraries, makes a network connection outside of OpenSearch), it should be in a separate repository. Please consider creating it in a standalone repository since each processor should be thought of like a *NIX command with input and output connected by pipes (i.e. a Search Pipeline). Each processor should do one thing and do it well. Otherwise, it could go into the OpenSearch repository under org.opensearch.search.pipeline.common. If you have doubts, just create an issue in OpenSearch Core and, if you have one, a new PR. Maintainers will help guide you.

History

This repository has also been used for discussion and ideas around search relevance. These discussions still exist here, however due to the relatively new standard of having one repo per plugin in OpenSearch and our implementations beginning to make it into the OpenSearch build, we have two repositories now. This repository will develop into a plugin that will allow OpenSearch users to rewrite search queries, rerank results, and log data about those actions. The other repository, dashboards-search-relevance, is where we will build front-end tooling to help relevance engineers and business users tune results.

Project Resources

Code of Conduct

This project has adopted the Amazon Open Source Code of Conduct. For more information see the Code of Conduct FAQ, or contact opensource-codeofconduct@amazon.com with any additional questions or comments.

License

This project is licensed under the Apache v2.0 License.

Copyright

Copyright OpenSearch Contributors. See NOTICE for details.

About

Search Request Processor: pipeline for transformation of queries and results inline with a search request.

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Java 90.7%
  • Shell 9.3%