Entity resolution: Running Entity Resolution

Once at least one mapping is configured, Admins and Source Managers can launch the Entity Resolution process from the Running tab.

This process scans your dataset to identify and group duplicate or related records based on the mappings defined.

Accessing the "Running" tab

From the Entity Resolution menu, click on the Running tab to manage the Entity Resolution process and monitor results.

From there, you are able to:

• Run the Entity Resolution process (if mappings are present)
• See Metrics of the last run (if at least one run has occurred)
• Clear Entity resolution results (if at least one run has occurred)

Starting the process

Click the Run button to begin the resolution process.

• The button switches to Stop
• A progress bar appears, showing the approximate time left
• Some metrics appear, showing the number of ingested records

ℹ️ The Entity Resolution process works incrementally, like data-source indexing. This means that :

• on the first run, all records from node categories listed in the mappings will be ingested and resolved.
• on the next run (or resume after a stop), only records that are new to the Entity Resolution engine since the last run will be ingested and resolved (i.e., new or modified nodes).

Stopping the process

At any time while Entity Resolution is running, click the Stop button to halt the process.

• The system transitions to a "Stopping" state.
• Once stopped, the Stop button turns into Resume

ℹ️ Stopping may take some time as the process is resolving records by batches, and waits for a batch to be resolved before completely stopping.

Resuming the process

If the process is stopped manually or interrupted due to an error, click Resume to restart it. The process will resume from where it left off.

Completion and success

When the process completes successfully:

• A green status message appears
• The Run button becomes active again
• Metrics are updated and fully displayed (see next section)

Handling errors

If the Entity Resolution process encounters an error:

• A banner is shown
• The Stop button becomes Resume
• Metrics display partial results gathered before the failure
• A See details button is shown for further technical information

You can resume the process once the problem has been fixed.

Metrics and process results

After the first successful run, key metrics become visible in the Running tab. These include:

Process metrics

• Last run Date/Time (in YYYY-MM-DD or hh:mm:ss)
• Duration of last process (in hh:mm:ss)

License usage

• License Consumption: The percentage of credits used on the license, based on records ingested.

  ℹ️Admin users can use the Manage License button to access more details about license consumption.

• Records Ingested: The total number of records ingested by Entity Resolution on this data-source.

Graph metrics

• Entities Created: Total number of resolved entities with more than one discovered relationship.
• Duplicates: Number of records identified as duplicates records, and which resolve to a single entity.
• Possible Duplicates: Number of POSSIBLY_SAME relationship.
• Possible Relationships: Number of POSSIBLY_RELATED relationship.

Clearing results

You may need to clear results during testing or after editing mappings.

ℹ️This will remove all nodes and relationships created by Entity Resolution from your dataset, and reset your license credits to start over.

Click the Clear results button and confirm your action to clean your dataset from Entity Resolution results.

ℹ️ The process may take some time depending on the number of resolved entities created by Entity Resolution.

⚠️ If you cancel the confirmation popup, no clearance is performed.

Upon success:

• Metrics are reset to zero
• License credits are fully reset
• Graph elements (resolved entities and edges) are removed from the graph