Understanding linked and standalone projects
Understanding linked processes
You can link an IBM Task Mining project to an existing parent IBM Process Mining process to analyze
the business activities that are present in it. IBM Task Mining is used to collect
user actions. In the IBM Task Mining project settings, you can associate each action to a business activity from the linked
IBM Process Mining project. If
you select multiple business activities, IBM Task Mining generates a IBM Process Mining project for each
business activity with the naming convention: Task Mining:
<business_activity_name>. You can then navigate from the parent IBM Process Mining project to each
IBM Process Mining project by
selecting the business activity and clicking the Launch Task Mining project button.
After the first recording with the Agent, you can see the Tasks in the Task
classification tab. In the Project Configuration Tab, you can assign each
task to a IBM Process Mining
business activity that is collected from the linked process.
In the following example, Ticketing contains these activities:
- Open ticket
- Wait user information
- Manage GitHub Issue
- Resolve call
- Update CRM
Each IBM Task Mining task in Ticketing is assigned to one or several of these business activities as shown in the following diagram.
Task Mining Project example 1
- Task Mining project name: Ticketing
- Organization: Task Mining Org
- Linked process: Ticketing
- Standalone: No
- Enable Obfuscation & Anonymization: if required
Table 1. Each IBM Task Mining task in Ticketing is assigned to one or several business activities
| Task | Business Activity |
|---|---|
| Open Issue | Manage Github Issue |
| Add Label | Manage Github Issue |
| Add Assignee | Manage Github Issue |
| Close Issue | Manage Github Issue |
| Create client | Update CRM |
| Update client | Update CRM |
| Create action | Update CRM |
IBM Task Mining automatically creates the following IBM Process Mining projects:
Table 2. Automatically created process mining projects
| Type | Name |
|---|---|
| Process Mining Project | Task Mining: Mange Github Issue |
| Process Mining Project | Task Mining: Update CRM |
From Ticketing IBM Process Mining project, you can select these two activities and drill-down into the linked IBM Task Mining projects by using the Task-mining button.
Understanding a stand-alone project
You can create a project in IBM Task Mining without associating it with any IBM Process Mining process. In this
case, an additional process is created in IBM Process Mining in
Standalone Task Mining:
- The process has the IBM Task Mining project name.
- The process contains all the related business activities as events.
- The Process ID is used to define cases (all the business activities need to share the process ID).
In the following example, there is no pre-existing Ticketing
IBM Process Mining project.
Therefore, IBM Task Mining does
not have any logs that are collected from IT systems.
Task Mining Project example 2
- Task Mining project name: Ticketing
- Organization: Task Mining Org
- Linked process:
- Standalone: Yes
- Enable Obfuscation & Anonymization: if required
Table 3. Tasks assigned to business activities
| Task | Business Activity |
|---|---|
| Open Issue | Manage Github Issue |
| Add Label | Manage Github Issue |
| Add Assignee | Manage Github Issue |
| Close Issue | Manage Github Issue |
| Create client | Update CRM |
| Update client | Update CRM |
| Create action | Update CRM |
IBM Task Mining creates the following IBM Process Mining projects:
Table 4. Automatically created process mining projects
| Type | Name |
|---|---|
| Process Mining Project | Ticketing |
| Process Mining Project | Task Mining: Mange Github Issue |
| Process Mining Project | Task Mining: Update CRM |
Apart from the recorded tasks and the contextual data, the processes on IBM Process Mining generated from IBM Task Mining contain the following information:
- CASEPRODUCTIVE
- The time that is spent by the user on the pages and windows that are associated with the case during the execution of the case (for example, by pressing keys or mouse buttons).
- CASEIDLE
- The time in which the user is not active on the pages and windows that are associated with the case during the execution of the case.
- APPLICATION
- Indicates the application on which the task is run.
- COUNTCOPY
- The number of copy actions performed within the case.
- COUNTPASTE
- The number of paste actions performed within the case.