Explore the Basics

Once Ocelescope is running, open http://localhost:3000 and load a log to start exploring.

Load a log

1. Open the upload dialog

   Use the upload action in the app to open the upload dialog.

2. Pick a log

   You have two options:

   • Use a built-in example. Ocelescope ships with a small set of default OCELs. Select one from the list to import it instantly, which is the fastest way to look around without preparing your own data.
   • Upload your own file. Drag an OCEL file onto the dropzone. Supported formats are .jsonocel / .json, .xmlocel / .xml, and .sqlite (OCEL 2.0). Traditional XES event logs can also be imported and are converted to an OCEL on the fly.

The same dropzone also accepts plugins and resources, so you can drop an OCEL, a plugin, or a saved resource in the same place.

Inspect the log

After import, the Log Overview gives you a first read on the data so you can confirm it loaded as expected before running any analysis. It summarizes:

• the number of events and objects,
• the object types and event types (activities) present,
• the attributes attached to events and objects, including value types and ranges,
• the event-to-object and object-to-object relations.

For a deeper, interactive look at how individual events and objects connect, open the Ocelot module, an Ocelescope adaptation of the original Ocelot tool. It adds four pages:

• Event Overview and Object Overview give an aggregated, per-type view of the events or objects in the log.
• Events and Objects let you browse individual records, search and filter by type, and inspect their attributes alongside the entities they relate to.

From the Events or Objects pages you can also open a graph visualization that lays out how the selected events and objects connect to each other.

Filtering

The Filter section lets you narrow a log down before analyzing it; and it offers visual support to do so. Get a slice of the log:

• See the activities/object types and their frequencies and choose the ones you want to keep.
• Zoom into everything that occurs within a specific timeframe
• See all attribute and type combinations and adapt them to include only those with values in specific ranges. Use the filters to search for specific ones.
• See all e2o and o2o relations and include only events/objects of the type with a specific number of these relations. Use the filters to search for relations you are interested in.

Once you apply a filter, the log gets a small blue filter icon in the log selection dropdown above. Every other view in Ocelescope (Log Overview, Ocelot, Discovery, plugins) then works on the filtered version of that log, until you remove the filter. You can also stack the filters.

Discovery

The Discovery section computes process models on the log you currently have selected. Out of the box it discovers:

• Object-centric Petri nets, and
• Directly-follows graphs (DFGs) During your discovery, you can apply ad-hoc filters scoped to that single computation: activity, object type, and frequency filters that drop infrequent event or object types, without touching the underlying log itself.

Discovered models can be downloaded as PNG, PDF, or SVG images directly from the Discovery view.

Resources and export

The resource view lists the OCELs and resources in your session. From there, each resource can be deleted, renamed, inspected, and exported.

• A log (filtered or not) can be exported as any of the supported OCEL 2.0 formats (.jsonocel, .xmlocel, .sqlite), or flattened to a single object type and exported as XES.
• Other resources, such as discovered models, can be exported as .ocelescope JSON files. These can be reloaded into Ocelescope later through the same upload dialog used for logs.

All resources, whether imported, filtered, or discovered, can also be used as inputs for plugin or module methods.

Next steps

Apply more process mining techniques via plugins.