Tamr can accelerate the schema mapping process by learning which input attributes are likely candidates to be mapped to unified attributes. The process works as follows. Some parts of this process are run by you and others are run by Tamr machine learning.
- Create unified attributes and partially map them, either manually or using bootstrapping (done by you).
- Update the unified dataset by running the job (initiated by you, completed by Tamr).
- Learn from mappings by running the job (done by Tamr).
- Generate mapping suggestions (initiated by you, completed by Tamr).
To control the number and quality of the suggestions that Tamr makes, you can specify how similar you want the mapping suggestions to be.
- Review and accept or reject Tamr's mapping suggestions (done by you).
- Repeat if necessary (done by you).
Steps 1 and 5 involve human feedback and input.
Steps 2-4 represent jobs that Tamr runs to learn and generate suggested mappings for attributes in the unified dataset.
You can also add more datasets to the project and repeat this process by running steps 2-4 again, and reviewing, accepting, or rejecting Tamr suggestions.
Updated 3 months ago
If you are running a mastering or categorization project, you can continue optimizing the schema for the unified dataset as needed. If, on the other hand, you are satisfied with the set of unified attributes, you can proceed to preview the records in the unified dataset.
|Schema Specifics for Mastering|
|Schema Specifics for Categorization|
|Previewing the Unified Dataset|