Entity resolution
Unify the variations of the same company's name under one common name as inconsistent data can cause problems when analyzing data or moving between systems.
What if you have transactional data that does not have a common ID? What if different systems do not have the same naming convention for the same account?
When analyzing data or moving it between systems, inconsistent data can cause problems. This blueprint helps you unify the variations of the same company's name under one common name.
The Entity resolution blueprint requires the following data sets:
Transactions (Data): The transactional data provided by the company. The same company names are recorded in multiple ways.
Account Lookup (Data): The "answer key" where each row represents a unique account ID with no missing or incorrect information.
Workflow
The same company names are recorded in multiple ways in Transactions (Data). For instance, "Mayo Clinic" is recorded as "Mayo Clinic" and "Mayo Clinic Hospital".
The Smart Matcher tool trains a model to map records from Transactions (Data) to a single account in Account Lookup (Data) , the answer key, and unifies the same company names under one name.
The Smart Matcher also adds a probability column. A higher probability indicates a better chance that the rows match correctly.
Important
When you created this Blueprint, you trained the model using the default data already in the pipe. If you want to use your own data, go to the Run tab and then select new sources for the pipe.
Once you have reviewed the results, you can export matched records for downstream processing. Export
Unmatched records can also be exported for further review and action. Export For Further Review (Export)