Page History
Versions Compared
compared with
Key
- This line was added.
- This line was removed.
- Formatting was changed.
Overview
This best practices document describes how to use Machine Learning keywords in curating collection and service records in the Common Metadata Repository (CMR) to improve metadata quality and discoverability of machine learning data and models.
Machine Learning Modelkeywords are a representation of predictive models that, when trained on a set of data containing certain features, enables a computer to identify similar features in other data. Machine Learning Training Data keywords are a representation of input data necessary for running a machine learning model.
Best Practices
Science keywords from the GCMD Keyword Management System (KMS) are important for the precise search and retrieval of data and should accurately represent the dataset data being described. At a minimum, one science keyword hierarchy must be provided, and this hierarchy the level of the keywords must go down to the 'Term' level of detail.. Additional keywords can be requested through the GCMD Keywords Community Forum.
- The Earth Science keywords should must be picked from the GCMD KMS.
- Examples
- OCEANS > MARINE ENVIRONMENT MONITORING > MARINE SURFACE ELEMENTS > MARINE SURFACE DEBRIS
- OCEANS > OCEAN CIRCULATION > DIFFUSION
- BIOSPHERE > ECOLOGICAL DYNAMICS > COMMUNITY DYNAMICS > BIODIVERSITY FUNCTIONS
- The full list of valid keywords can be found
- here.
- Examples
- The Machine Learning Training Data keywords should must be picked from the GCMD KMS.
- Examples
- MACHINE LEARNING TRAINING DATA > LABELS > RASTER LABEL
- MACHINE LEARNING TRAINING DATA > SOURCE > VECTOR SOURCE
- MACHINE LEARNING TRAINING DATA > SOURCE > RASTER SOURCE
- The full list of valid keywords can be found
- here.
- Examples
- The Machine Learning Model keywords should must be picked from the GCMD Keyword Management System. The KMS.
- Examples
- MODELS > MACHINE LEARNING MODELS > SUPERVISED
- MODELS > MACHINE LEARNING MODELS > DECISION TREE > ISOLATION FOREST
- MODELS > MACHINE LEARNING MODELS > DEEP LEARNING > GENERATIVE ADVERSARIAL NETWORKS
- The full list of keywords can be found
- here.
- Examples
Machine Learning Training Data (Model Training Data)
:- Describe your machine learning Machine Learning training data as a UMM Collections Collection (UMM-C) compliant collection record in CMR.
- See the full list of UMM-C fields at UMM-C Schema Documentation.
- Add relevant ' Machine Learning Training Data ' keywords describing the input data necessary for running a machine learning model model.
- Add relevant ' Earth Science ' keywords describing what is being measured as part of the training data set set.
- Ingest your UMM-C record into the CMR. (Note: A NASA Agency AUID and Launchpad authentication is required for metadata ingest operations.)
- Add the CMR Tag as 'machine.learning' Groups could apply the
- CMR tag to their records if given the proper permissions and approvals by ESDIS
- your record. (Note: Tag permission is granted through the provider via the INGEST_MANAGMENT_ACL. Users can add tags if they are granted the appropriate permission.)
- To add a tag to a collection record, see the Search API Documentation.
Machine Learning Models (Model)
:- Describe your machine learning model as a UMM Services (UMM-S) compliant record in CMR.
- Add relevant ' Machine Learning Model ' keywords describing the type of model that was used to train the data.
- Ingest your UMM-S record into the CMR. (Note: A NASA Agency AUID and Launchpad authentication is required for metadata ingest operations.)
- Add Create a collection association to relevant Machine Learning Training Data collections (from above) in the CMRThis can be done via MMT or the CMR API. (Note: Collection association is granted through the provider via the INGEST_MANAGMENT_ACL. Users can add tags if they are granted the appropriate permission.)
- To associate a service with one or more collections using the Metadata Management Tool (MMT), see the MMT User Guide.
- To associate a service with one or more collections using the CMR API, see the CMR Search API Documentation.
Example Collection Record
References
Overview
Content Tools