Process Learning Recommendations
This process expands and reconciles Oracle Learning recommendation initiatives separately from required or voluntary learning initiatives. It evaluates the defined audience to identify the affected set of workers. Then it creates or withdraws recommendations depending on the intended audiences.
The process automatically generates child processes to handle the necessary work.
You've two options for execution type:
-
Reconcile: The process evaluates the target audiences and assigns, withdraws, or reactivates recommendations as appropriate. We recommend that you schedule the process to run with this option on a daily basis, 1 hour after the Process User Access process runs
If you use required or voluntary learning initiatives, keep the Evaluate Person IDs for Assignment Rule and Reconcile Dynamic Assignments processes on their appropriate schedule. Reconciliation still requires them.
- Expansion: The process creates the recommendations for the target audience when learning administrators create recommendations. You don't need to scheduled the process with this option unless you deselect Process recommendations immediately after user saves changes on the Configure Background Processing page. If you deselect the parameter, you need to run the process with this option to reconcile learning initiatives.
Leave the Initiative Number parameter blank for normal operations. Enter a specific initiative number when you want to expand or reconcile the recommendations for just that initiative.
Change Execution Details
To change these defaults, go to :
- The delivered run as default for expansion and reconciliation processes is creator, but you can select a specific user. For example, to avoid learning initiative issues that can result when creators leave, select a generic user account that persists indefinitely.
- To let learning administrators override the default, select Override Default Expansion and Reconciliation - Run as User.
-
To reduce the overall time spent expanding and reconciling learning recommendations, change the concurrency details related to parallel processing and when to process recommendations.
Concurrency Option Description Maximum Number of Parallel Bulk Scheduled Jobs Specify how many analyses to evaluate in parallel for all recommendations and recommendation initiatives that need expansion or reconciliation based on selected execution type. The value also decides how many recommendations to reconcile in parallel.
The supported values are 1 through 200, and the default value is 1. We recommended you set the parameter to at least 5. You can then increase as appropriate to reduce the overall time it takes the process to complete.
Maximum Number of Parallel Threads per Bulk Scheduled Job Specify the maximum threads to use per recommendation or recommendation initiative during reconciliation, to create or update recommendation records.
The supported values are 1 through 20, and the default value is 1. We recommended that you keep the value at 1. The reason is that the maximum threads calculation multiplies this value with the Maximum Number of Parallel Threads per Ad Hoc UI Jobs value. Consider all other background processes used in Oracle Learning before increasing these values.
Process recommendations immediately after user saves changes When selected (default), the process is submitted automatically when a learning administrator creates, edits, or cancels a recommendation or recommendation initiative.
When unselected, you need to schedule this process as recurring in Expansion mode to expand recommendations, and in Reconcile mode to reconcile access. This option lets you schedule expansion and reconciliation during off peak periods.
Maximum Number of Parallel Ad Hoc UI jobs Specify how many analyses to evaluate in parallel for the newly created recommendation or recommendation initiatives.
The supported values are 1 through 200, and the default value is 1. We recommended you set the parameter to at least 2. You can then increase as appropriate to reduce the overall time it takes the process to complete.
Maximum Number of Parallel Threads per Ad Hoc UI Jobs Specify the maximum number of threads to use to create the recommendation records needed for the expansion process.
The supported values are 1 through 20, and the default value is 1. We recommended that you set the value to 2.