Apromore allows us to discover a process map or a BPMN model from an event log. A process map (a.k.a. directly-follows graph) is a visual representation of the log as a graph where nodes capture process activities and directed arcs between them capture sequential order relations between the activities. For example, an arc going from activity Accept order to activity Check order indicates that in the log, we can observe that process cases flow from Accept order to Check order. Process maps are a simple yet effective means to understand the basic order of relationships between process activities. As such, they are the most common type of model discovered by commercial process mining tools.
View process map
To view the process map, double-click on a log. Alternatively, we can right-click the log Discover model > From full log .
A window will open up, showing the process map discovered from the log.
We can adjust the complexity of the discovered map by increasing or decreasing the frequency or duration of nodes and arcs visualized in our process map. The default values for the nodes and arcs sliders are 100% and 10%, respectively.
If a log has more than 250 nodes, only the most frequent 2500 nodes are displayed in the process map or BPMN model in Process Discoverer.
We can abstract a process map by Case frequency or Average Duration. For example, if we abstract by case frequency and shift the arcs or nodes slider towards the left, more edges/nodes with low case frequency will be removed from the process map.
Similarly, if we abstract by Average Duration and shift the arcs or nodes slider towards the left, more edges/nodes with low average duration will be removed from the process map.
We recommend you to use the abstraction sliders as follows.
First, use the Nodes slider to select the percentage of nodes you wish to display in the canvas. For example, if you set the Nodes slider to 80%, Apromore will display only the 80% most frequent nodes and remove the 20% most infrequent.
Next, you may adjust the number of arcs displayed on the canvas. If you set the Arcs slider to 100%, then Apromore will display all the arcs between the current set of nodes. As you move the Arcs slider down from 100% down to 0%, Apromore will hide the most frequent arcs. Importantly, when Apromore hides an arc, it will always ensure the following two properties:
The percentage of displayed nodes does not change. In other words, as you move the Arcs slider down from 100 to 0%, no node will be hidden along the way. If the Nodes slider is set to 80%, you will always see the 80% most frequent nodes, regardless of the position of the Arcs slider. For example, if the Nodes slider is set to 80% and the Arcs slider is set to 10%, Apromore will display 80% of the nodes, and a small number of arcs between these nodes.
Every node has at least one incoming arc and one outgoing arc. In other words, there will be no disconnected nodes or disconnected fragments. This property is important. It ensures that if you switch the current view to the BPMN view, the BPMN model will be fully connected.
Because of the above two requirements, some arcs cannot be hidden. For example, it may happen that a node has many incoming arcs, each of which is relatively infrequent. As you move the slider from 100% to 0%, these infrequent arcs will be hidden. At some point, it may be that only one of the incoming arcs is displayed. This last remaining arc will not be deleted because otherwise the node would be disconnected from the rest.
In fact, Apromore’s arcs slider is designed to ensure a stronger property: Every node will always be on a path from the start node to the end node. This property is required in order to transform the process map into a correct BPMN model.
You can also use the abstraction sliders to remove the most infrequent nodes and the most infrequent arcs, instead of removing the most frequent nodes and arcs. To remove the most infrequent arcs, you should click the invert order toggle icon . When the invert order to toggle is on, Apromore will:
Hide the most infrequent nodes when you move the Nodes slider down from 100% to 0%
Hide the most infrequent arcs when you move the Arcs slider from 100% to 0%
Similarly, you can use the abstraction sliders to remove the nodes or arcs with the highest duration. To achieve this behavior, select Average duration in the menu Abstract by located above the sliders. When the Average duration option is selected, Apromore will hide the nodes or arcs with highest average duration (the slowest arcs) when you move the corresponding slider towards 0%. Conversely, if the invert order toggle is in the On state and the option Average duration is selected, Apromore will hide the nodes or arcs with lowest average duration (the faster arcs) when you move the corresponding slider towards 0%.
Due to limitations in screen sizes and resolution, Apromore will never display more than 250 arcs in total. If the number of activities or the number of resources exceeds this threshold, Apromore will display the most frequent 250 nodes and ignore the remaining ones.
We can also check more detailed statistics.
View case inspector
To inspect individual cases, use the Case Inspector. To open the Case Inspector, click on the number of cases under the Log statics section.
The Case Inspector window will open.
Click on a specific case to visualize it exclusively in the Process Discoverer.
To check the length of the cases, click on the Activity instances column. Cases will be sorted from the shortest to the longest.
We can also sort them by the duration of the case. We do this by clicking the Duration column.
We can sort the case inspector table in ascending or descending order of a column by clicking the column header multiple times.
To download the list of cases and their statistics, click on Download.
View case timeline
We can also display a case timeline. A case timeline is a chart that displays the timeline of activities as a Gantt chart. To display the case timeline, double-click a case in the case inspector.
This displays the case timeline for the selected case. This contains the case timeline chart and the case data from the log.
By default, the case timeline includes the waiting times in the case. We can, however, remove the waiting times by checking “Remove waiting time”. Now, the chart is displayed, ignoring waiting times between activities.
We can also display the case timeline for only specific activities. In the case data, select the activities we wish to display.
To display the case data alone, click the inverted triangle button right above the slider. You can also drag it to adjust the height of the case data dynamically.
By default, Apromore displays the case timeline chart with the minimum width. We can however increase this chart width by entering our preferred width and clicking OK.
Finally, we can download the case timeline or the case data.
Click the download button in the top-right corner.
Click Download timeline as PNG to download the chart or Download case data as CSV to download the case data.
The respective file is zipped and downloaded.
View activity inspector
We can also inspect an individual activity in Process Discoverer. To inspect individual activity, use the Activity Inspector. To open the Activity Inspector, click on the number of activities under the Activities chart in Log statics section.
The Activity Inspector window will open.
To check the length of the activity, click on the Activity instances column. To check activity variant frequency, click on Percentage(%) column.
To download the list of activities and their statistics, click Download.
To check the individual activity statistics, we can simply click on the activity. We can also select multiple activities at once by pressing Ctrl (for Windows) or Command (for Mac) and dragging the mouse over the part of the model which consists of these activities.
Visualize by Frequency
We can also view the event log’s simple statistics, such as the total, median, minimum, maximum, and average number of times an activity is executed. This information is provided as a label on the activities/arcs, color of activities, and arcs’ thickness. The darker the blue color, the higher the number of times that activity has been observed in the log. The thicker arc, the higher the frequency of that arc. We can use the Frequency drop-down list in the Visualization settings section to view different statistics.
Visualize by Duration
Additionally, we can view the statistics on the time performance of the activities and arcs in the process map using the Duration drop-down list in the Visualization settings section. These are total, mean, median, minimum, maximum, and average duration of each arc (indicating the waiting time before starting a given activity, once the previous one has been completed), and total, mean, median, minimum, maximum, and average duration of each arc duration of an activity (a.k.a. the activity’s processing time). Suppose the log only has completion timestamps for each activity and not their start timestamp. In that case, these performance statistics will combine both processing time and waiting time into a single time statistics visualized on the arc. At the same time, activities will be shown as having an instantaneous duration.
Like frequency statistics, we can also visualize time performance statistics via labels on activities and arcs and via colors and line thickness (on a red scale) for activities and arcs.
Visualize by Frequency and Duration
We can also view both Frequency and Duration at the same time by clicking the . For example, if we are viewing our process map in Frequency mode, we can also view the Duration as a secondary metric by clicking on the .
Visualize by Cost
Apromore allows displaying cost metrics on top of a process map or a BPMN model in the Process Discoverer.
The Cost overlay functionality can be used only with the logs having the Roles column present while uploading the event log to Apromore.
To check the cost metrics, click on Manage cost centers button placed right above the Abstraction Settings panel.
The Cost center management window appears, consisting of Roles, Cost per hour and a Calendar. A Currency can be defined below.
Enter the cost values next to each Role
We can also assign calendars to each role. For instance, other roles may work Mondays to Fridays while the “Supplier” work Tuesdays to Fridays. Their working schedule affects the costs so we can create and assign respective calendar to roles. To assign a calendar, click the assign calendar button.
This brings the calendar management window which displays the list of available calendars. Click on a calendar and click Apply calendar to assign it to the role.
We can do this for all the roles.
Before assigning a calendar to a role, the calendar must first be created. To learn more about creating calendars, see Manage calendars
We can also choose a currency from the Currency drop-down and click Apply.
To see how the Resource costs are distributed among different process activities, click on the Cost overlay available under the Visualization settings section, Overlay subsection placed top left in the Process Discoverer.
Visualize by Different Perspectives
Visualizing the handover between activities is not the only way a process can be analyzed. There will be times when we may be interested in assessing if a specific resource or group of resources are overloaded with work. When clicking on Perspective we can decide which attribute of the log will focus on the process map.
For example, to visualize the handover of work among a group of resources, we can select the Resources option. This option will map each actor’s organizational role in our process to a node and connect two nodes if a handover of work occurs between the two nodes.
Display Rework Metrics
Rework is a common type of waste in a business process. The occurrence of rework increases the cycle time and cost of cases in a process.
To help identify and quantify rework in a process, we can display rework metrics in Process Discoverer. There are three kinds of rework metrics in Apromore:
Total rework: Total number of times an activity has been repeated after having been executed once in a case.
Rework duration: Total duration of the reworked activity instances.
Rework cost: Total cost of the reworked activity instances.
To display rework metrics in the Process Discoverer, open an event log from the Portal. In the Overlay section, click Frequency to specify working in the Frequency overlay. Click the dropdown, and select Total rework.
The process map now displays the total rework for each task.
We can also display the total duration the resources spend on the rework. Click Duration to switch to the duration overlay and select Total rework.
The process map displays the duration of the total rework for each task.
Similarly, we can display the total rework cost for each task. To display the cost of the total rework, the event log must have a Role attribute, and a cost table must be defined for this event log.
In the Overlay section, click Cost to switch to the cost overlay. Select Total rework from the dropdown.
The process map displays the cost of the total rework for each task.
Filter Log in Process Discoverer
When analyzing an event log, we may be interested in isolating a particular type of behavior or removing a specific activity. An event log can be filtered by clicking on the Filter icon and creating a new filter that fits our needs. For more information on the optimal use of Filter, we suggest viewing the Filter Log manual.
View BPMN Model
Whenever the insights deriving from the analysis of a process map are not sufficient, the same functionalities are offered on top of a BPMN model. Changing the view from Process Map to BPMN model will automatically discover a BPMN model from an event log. When visualizing a BPMN model, the slider Parallelism offers the possibility to adjust the amount of parallelism (e.g., AND and OR gateways) discovered by the plugin.
Export Process Map/BPMN Model
Apromore allows us to save or download a log/BPMN model from Process Discoverer.
Save Filtered Log
To save a filtered log, click the save button.
Click Save filtered log.
We are then prompted to enter the name of the log and a short description.
Click OK to save the filtered log. The filtered log will be saved in the same Apromore Workspace directory as the opened log.
Save BPMN Model
A BPMN model can also be saved from Process Discoverer. Click the save icon (first ribbon) in the top left corner. Click Save BPMN model.
To save a BPMN model, the BPMN model view must be toggled on.
We will be prompted to enter the name of the model, a short description and the version number. Additionally, we can specify the folder where the BPMN model will be stored. We can also specify whether the BPMN model should include the simulation parameters.
Click OK to save.
Download PDF, PNG and JSON
Apromore allows us to download a BPMN model or filtered log in different file formats. We can download the model or log in PDF, PNG or JSON formats. To download a model/log, click the download icon in the top left ribbon.
Specify the file format in which the model/log to be downloaded.
The file is then downloaded to our local machine in the specified format.
We can also search for an activity by using the search bar on the top-right.
For ease of view, the search results get highlighted in the process map.