Conformance Analysis: Difference between revisions

From QPR ProcessAnalyzer Wiki
Jump to navigation Jump to search
(404307)
 
(35 intermediate revisions by 2 users not shown)
Line 1: Line 1:
The Conformance Analysis View incorporates the [[Conformance Checking in QPR ProcessAnalyzer|conformance checking]] capability of QPR ProcessAnalyzer. There are two views in the conformance analysis: the other shows the violations, and the other shows possible root causes for the nonconformance using the Influence Analysis for Case Attributes.
The Conformance Analysis dashboard incorporates the [[Conformance Checking|conformance checking]] capability of QPR ProcessAnalyzer. There are two ready-made dashboards in the conformance analysis: the other shows the violations, and the other shows possible root causes for the nonconformance using the Root causes analysis. For customized needs, you can also create your own dashboards that contain all the conformance analysis elements available in the ready-made dashboards.


== AS-IS Model ==
== Editing Design Model ==
This panel shows the real model data. You can use the AS-IS Model to filter cases. Note that you are not bound to using just the Flowchart analysis, as you can change the analysis by right-clicking the analysis and selecting a different type of analysis shown on the AS-IS Model panel.
The design model describes precisely how the process should go using [https://www.omg.org/spec/BPMN/2.0/ BPMN 2.0] notation. The design model editor can be opened by clicking the '''Edit Design Model''' button. When editing the design model, only a subset of the BPMN elements are supported which are following: Start event, End event, Task, Gateway and Connector. Unsupported elements are not available in the tool palette.


== Conforming vs Nonconforming Cases ==
It's recommended to create the design model automatically based on the data in the as-is model. To do this, select from the context menu either '''Auto-create tasks and flows''' or '''Auto-create tasks only''':
This panel shows summary information about the size of conforming and nonconforming cases.
* '''Auto-create tasks and flows''': Creates a design model that is effectively the same as the as-is model. Thus, the as-is model will conform perfectly to the design model. After the tasks and flows are created, you can edit the design model and remove the events and flows that shouldn't occur in the process. Note however, that the amount of flows that is created can easily be overwhelming. To overcome this, you can first use the flowchart to filter out those events and flows that shouldn't occur in the process, then use the auto-create option. Then, after creating the design model, you can remove the filter to check the conformance of the whole model.
* '''Auto-create tasks only''': Generates only the tasks without any flows. The result is a model that needs to be filled with flows to be a valid design model.


== Conformance Trend ==
Note that the previous design model is replaced by the generated model.
This panel shows the case counts of conforming cases and total cases per month.


== Top Violations ==
== Conforming vs Nonconforming Cases ==
This panel shows the violations against the design model:
The [[QPR ProcessAnalyzer BPMN Editor#Conformance_Statistics|Conformance Statistics]] component shows summary information about the size of conforming and nonconforming cases.
* '''Cases''': the percentage of cases having the violation.
* '''Duration Impact''': shows the impact the violation has on case duration, i.e. the average duration of the total amount of cases in the model compared with the average duration of the violating cases. For example, the Duration Impact might show that, on average, cases with the violation take 29 days longer to finish.
* '''Steps Impact''': shows the impact the violation has on event count, i.e. the average event count in the total amount of cases in the model compared with the average event count in the violating cases. For example, the Steps Impact might show that, on average, there are 1.5 less events in the cases with the violation.


The main types of violations are as follows:
== Conformance Trend Chart ==
* '''Undesired event''': the event exists in the model, but not in the design model.
This chart shows the conforming cases count and total cases across time, which visualizes how the conformance rate has developed over time.
* '''X occurred directly after Y''': an event occurred in the wrong sequence in the model, i.e. according to the design model, the event X should not occur right after event Y.
* '''Case did not complete''': case finishes in the model data, but according to the design model, the case should still continue.


== Top Violating Variations ==
== Top Deviations (in-memory) / Deviating Flows (Snowflake) ==
Shows the variations that violate the design model rules sorted by size.
This table shows specific reasons for the conformance deviation for the non conforming cases. For the in-memory models, all deviations are shown, and for the Snowflake models, only the flow related deviations are shown. There are following columns:
* '''Top Deviating''': this column shows the type of deviation (i.e., the actual reason for the nonconformance). The types of deviations are as follows:
** '''Undesired event''': the event occurred in the eventlog, but it's not in the design model and thus it should not occur at all. This deviation is not shown in Snowflake models.
** '''<X> occurred directly after <Y>''': an event occurred in the wrong sequence in the eventlog, i.e. according to the design model, the event X should not occur right after event Y.
** '''<X> is the starting event''': case starts with an event that is not allowed in the design model. This type of deviation is reported when the first event in of the case can occur later in the process but not as the first event. If the event is not allowed at all in the process, the deviation would be reported as ''undesired event''.
** '''<X> is the ending event''': case reaches its last event (thus far) in the eventlog, but according to the design model, the case should continue. Note that this "deviation" may occur simply because the continuation events haven't yet occurred, and thus it's not a real deviation in the process.
* '''Cases''': Percentage of cases having the deviation.
* '''Duration Impact''': Shows the impact the deviation has on the case duration, i.e. the average duration of all cases compared with the average duration of the deviating cases. For example, the Duration Impact might show that, on average, cases with the deviation take 5 days longer to finish than on average.
* '''Steps Impact''': Shows the impact the deviation has on the number of events, i.e. the average number of events among all cases compared with the average number of events in the deviating cases. For example, the Steps Impact might show that, on average, there are 1.5 events more in the cases with the deviation.


== Show Root Causes ==
== Undesired Events (Snowflake) ==
Clicking this button will open a view similar to the current one, except that it uses Influence Analysis for to show which factors influence the cases that violate the design model rules:
This table shows all event types that not conforming with the design model, i.e., they are event types appearing in the eventlog that are not in the design model at all. The table shows the duration and steps impact, similar to the ''Deviating Flows'' table.
* '''Cases #''': the tootal number of cases that have this case attribute value.
* '''Nonconformance %''': the percentage of nonconforming cases of the total number of cases shown on this row.
* '''Contribution %''': the percentage of the cases with the shown attribute value that contributes to the cases not conforming to the design model rules.


== Design Model ==
== Conformance Flowchart ==
The design model is the set and order of events against which the data in the QPR ProcessAnalyzer model is tested. The design model is defined using [https://www.omg.org/spec/BPMN/2.0/ BPMN 2.0] notation.  
For in-memory, the flowchart highlights the deviating event types and flows with red color. The deviations information is the same as in the [[#Top_Deviations|Top Deviations]] table with one exception: In the flowchart, for the deviating event types also the incoming and outgoing flows are shown as deviations . These kind of deviating flows are not listed in the ''Top Deviations'' table to make the table more readable because the listed undesired event already implies that the related flows are deviating.


=== Creating the Design Model ===
In the flowchart, note that only the visible event types and flows may be shown as deviations, so there may be deviating event types and flows that are not shown due to the [[Process_Flowchart#Visibility_settings|flowchart visibility settings]]. Note also that the conformance checking stops to the first deviating flow while checking each case, and thus the checking might not detect correctly possible deviations that are occurring after the first deviation. The first deviation can be "skipped" by either filtering the model or changing the design model.
It's recommended to create the design model automatically based on the data in the as-is model. To do this, click on '''Auto-create Design Model'''. You will then have two options for generating the design model automatically:
* '''Create Tasks and Flows''': this will create a design model that is effectively the same as the as-is model. Thus, the as-is model will conform perfectly to the design model. After the tasks and flows are created, you can edit the design model and remove the events and flows that shouldn't occur in the process. Note however, that the amount of flows that is created can easily be overwhelming.
* '''Create Tasks Only''': this will generate only the tasks without any flows. The result is a model that needs to be filled with flows to be a valid design model.
Note that the previous design model is replaced by the generated model.


=== Editing the Design Model ===
For Snowflake, the Conformance Flowchart shows the deviating event types, but not deviating flows. Still, flows where the starting or ending event type is deviating, also the flow is shown as red.
The design model editor is open after using either of the creation methods described above. When viewing the analysis, you can open the design model editor by clicking '''Edit Design Model'''.  


When editing the desgign model, use only the following elements: Start event, End event, Task, Gateway, and Connector.
== Top Deviating Variations ==
Shows the most common variations in the process that are not conformant with the design model.


== Show Root Causes for Deviations ==
Clicking the '''Show Root Causes''' button opens a view showing the Root causes analysis listing case attribute values that correlate with conformance violations. You can go back to the previous violations view by clicking the '''Show Violations''' button.


[[Category: QPR ProcessAnalyzer]]
[[Category: QPR ProcessAnalyzer]]
[[Category: Deployment Packages]]
[[Category: QPR UI]]

Latest revision as of 18:02, 1 June 2023

The Conformance Analysis dashboard incorporates the conformance checking capability of QPR ProcessAnalyzer. There are two ready-made dashboards in the conformance analysis: the other shows the violations, and the other shows possible root causes for the nonconformance using the Root causes analysis. For customized needs, you can also create your own dashboards that contain all the conformance analysis elements available in the ready-made dashboards.

Editing Design Model

The design model describes precisely how the process should go using BPMN 2.0 notation. The design model editor can be opened by clicking the Edit Design Model button. When editing the design model, only a subset of the BPMN elements are supported which are following: Start event, End event, Task, Gateway and Connector. Unsupported elements are not available in the tool palette.

It's recommended to create the design model automatically based on the data in the as-is model. To do this, select from the context menu either Auto-create tasks and flows or Auto-create tasks only:

  • Auto-create tasks and flows: Creates a design model that is effectively the same as the as-is model. Thus, the as-is model will conform perfectly to the design model. After the tasks and flows are created, you can edit the design model and remove the events and flows that shouldn't occur in the process. Note however, that the amount of flows that is created can easily be overwhelming. To overcome this, you can first use the flowchart to filter out those events and flows that shouldn't occur in the process, then use the auto-create option. Then, after creating the design model, you can remove the filter to check the conformance of the whole model.
  • Auto-create tasks only: Generates only the tasks without any flows. The result is a model that needs to be filled with flows to be a valid design model.

Note that the previous design model is replaced by the generated model.

Conforming vs Nonconforming Cases

The Conformance Statistics component shows summary information about the size of conforming and nonconforming cases.

Conformance Trend Chart

This chart shows the conforming cases count and total cases across time, which visualizes how the conformance rate has developed over time.

Top Deviations (in-memory) / Deviating Flows (Snowflake)

This table shows specific reasons for the conformance deviation for the non conforming cases. For the in-memory models, all deviations are shown, and for the Snowflake models, only the flow related deviations are shown. There are following columns:

  • Top Deviating: this column shows the type of deviation (i.e., the actual reason for the nonconformance). The types of deviations are as follows:
    • Undesired event: the event occurred in the eventlog, but it's not in the design model and thus it should not occur at all. This deviation is not shown in Snowflake models.
    • <X> occurred directly after <Y>: an event occurred in the wrong sequence in the eventlog, i.e. according to the design model, the event X should not occur right after event Y.
    • <X> is the starting event: case starts with an event that is not allowed in the design model. This type of deviation is reported when the first event in of the case can occur later in the process but not as the first event. If the event is not allowed at all in the process, the deviation would be reported as undesired event.
    • <X> is the ending event: case reaches its last event (thus far) in the eventlog, but according to the design model, the case should continue. Note that this "deviation" may occur simply because the continuation events haven't yet occurred, and thus it's not a real deviation in the process.
  • Cases: Percentage of cases having the deviation.
  • Duration Impact: Shows the impact the deviation has on the case duration, i.e. the average duration of all cases compared with the average duration of the deviating cases. For example, the Duration Impact might show that, on average, cases with the deviation take 5 days longer to finish than on average.
  • Steps Impact: Shows the impact the deviation has on the number of events, i.e. the average number of events among all cases compared with the average number of events in the deviating cases. For example, the Steps Impact might show that, on average, there are 1.5 events more in the cases with the deviation.

Undesired Events (Snowflake)

This table shows all event types that not conforming with the design model, i.e., they are event types appearing in the eventlog that are not in the design model at all. The table shows the duration and steps impact, similar to the Deviating Flows table.

Conformance Flowchart

For in-memory, the flowchart highlights the deviating event types and flows with red color. The deviations information is the same as in the Top Deviations table with one exception: In the flowchart, for the deviating event types also the incoming and outgoing flows are shown as deviations . These kind of deviating flows are not listed in the Top Deviations table to make the table more readable because the listed undesired event already implies that the related flows are deviating.

In the flowchart, note that only the visible event types and flows may be shown as deviations, so there may be deviating event types and flows that are not shown due to the flowchart visibility settings. Note also that the conformance checking stops to the first deviating flow while checking each case, and thus the checking might not detect correctly possible deviations that are occurring after the first deviation. The first deviation can be "skipped" by either filtering the model or changing the design model.

For Snowflake, the Conformance Flowchart shows the deviating event types, but not deviating flows. Still, flows where the starting or ending event type is deviating, also the flow is shown as red.

Top Deviating Variations

Shows the most common variations in the process that are not conformant with the design model.

Show Root Causes for Deviations

Clicking the Show Root Causes button opens a view showing the Root causes analysis listing case attribute values that correlate with conformance violations. You can go back to the previous violations view by clicking the Show Violations button.