Start Spend Classification
Use the Start Spend Classification scheduled process to begin the process of classifying spend data.
If the number of transactions available for classification is greater than 100,000 then multiple batches are automatically created to accommodate all the transactions, eliminating the need for creating multiple batches manually.
When to Use
Use this process periodically when you want to classify the different types of spend transactions. Submit the scheduled process to classify spend data, without having to navigate to the spend classification work area.
Privileges Required
- Administer Spend Classification Application (POI_ADMINISTER_SPEND_CLASSIFICATION_PRIV)
- Manage Spend Classification Batch (POI_MANAGE_SPEND_CLASSIFICATION_BATCH_PRIV)
- View Spend Classification Work Area (POI_SPEND_CLASSIFICATION_WORKAREA_PRIV)
Specifications
Review these specifications before you set up the scheduled process.
Specification | Description |
---|---|
Job Type | Ad hoc or scheduled. |
Frequency | Run as frequently as required by the business. Typically, businesses like to classify their spend transactions fortnightly, monthly, or quarterly as per business needs. |
Time of Day | Any time. |
Duration | Run time of this job varies considerably with the number of transactions available for classification. For example, if run every 2 weeks and the volume of invoices being classified is less than 100,000 then the run time is expected to be less than 60 minutes. For higher volumes, it may go up to a couple of hours. |
Compatibility | If using the same knowledge base, there should be only one instance of the job running at any one time. Even if you are using multiple knowledge bases, separate the schedules so that the processes don’t overlap each other. |
Parameters
Parameter | Optional or Required | Description | Parameter Values | Special Combinations Required | Notes |
---|---|---|---|---|---|
Data Set | Required | Group of unclassified or misclassified spend transactions that needs to be categorized or re-categorized using spend classification. | The parameter values are:
|
None | N/A |
Business Unit | Required if you set the Data Set parameter to a data set that is available (seeded data set) with the application. Otherwise, it’s optional. | Name of the business unit for which data needs to be classified. | Name of the business unit. | None | Required if data set is one of:
|
Knowledge Base | Required | The knowledge or patterns present in a data set. The knowledge base is created as a result of training the machine learning engine. | User-defined knowledge base with the status Complete. | None | N/A |
Batch Name | Optional | Name given to the group of transactions that is being classified. | Free text | None | If left blank, batch number, which is a unique identifier generated by the system will be assigned as batch name. |
Transactions From | Required | Transaction date from which the spend transactions should be classified. | Calendar | None | To consider all the transactions available for classification or while scheduling, set a date like 01-Jan-1900. |
Transactions Until | Optional | Transaction date till which the spend transactions should be classified. | Calendar | None | To consider all the transactions available for classification or while scheduling set a date like 01-Jan-2090. |
Apply Rules | Required | Indicates if rules should be applied while submitting the spend classification process. | Yes or No | None | The default value is No. |
Troubleshooting Information
- You can view the status of the process on the Scheduled Processes UI.
- The validation errors and warning messages that prevented the classification of categories is displayed in the error log.
- Any interactive warning validations aren't performed.
- When the program is submitted, you can choose options such as Resubmit, Put on Hold, Cancel Process, Release Process as provided by the scheduled processes UI.