Drag an object-centric event log to this drop zone ...
Please select the correct values for the separator and the quote character of the CSV. Moreover, please provide a mapping for the columns (i.e., which column is the activity, timestamp, object type, event attribute, event identifier).Separator: Quote characters: Object Quotes:
The page shows the list of events of the log. Each event is distinguished by its fundamental properties (ID, activity and timestamp), a list of related objects for each type, and values for some attributes. Clicking an object, it is possible to see the lifecycle of the object.
The page shows the properties of the objects, after the selection of an object type. Each object is distinguished by the timestamp of the first event of its lifecycle, the timestamp of the last event of its lifecycle, and the difference between them.
To start, please select an object type:
An execution describes the connection between the lifecycle of different objects. Executions can be grouped together if they are equivalent in the involved object types.
E agg. place
E agg. attr.
This page permits to explore, using the SQL language, the table of the features (events=OCEL_EVENTS, objects=OCEL_OBJECTS) on the provided log. Patterns such as excessive repetition of an activity during the lifecycle of an object can therefore be considered here.
The isolation forest calculated on the objects' features permits to identify the objects with peculiar behavior (different from the one of the other objects). Objects which are more anomalous are reported first.
The first two components of the discovered features are illustrated in this page. This permits to visually identify the groups of objects with similar behavior (in terms of the calculated features).
Select the dimensionality reduction technique:
A decision tree helps to understand the conditions which determine a given outcome of a case:
Select the categorical attribute:
Select the attributes to ignore:
Shows the correlation between a selected feature and all the others features. This permits to understand which other factors influence the given feature.
Select the target variable:
Select the type of statistic that should be considered:
Some conformance rules (the value of a feature should be contained in a range) are inferred automatically from the extracted features. Anomalous objects are identified by such rules.
|Minimum Normal Value
|Maximum Normal Value
|Number of Deviations
The page shows the objects of the given type which are related to events of the given activity in extraordinary number (lower than avg - stdev, or higher than avg + stdev). It is offered a possibility to filter on the events related to such objects, to understand the control flow in these situations.
The page shows the log skeleton constraints (default noise: 5%) that are violated by the objects of the given type. Each rule is reported with a description, the overall number of violations, and a possibility to filter the events related to such objects.
The page identifies deviations in the duration of the lifecycle of the objects. Objects having an exceptional durations are here identified, along with the number of deviations and a possibility to filter the events related to such objects.