Ebi

Ebi is a stochastic process mining software suite, maintained by the BPM group of RWTH University, Aachen, Germany. It contains several algorithms and techniques that perform analyses on event logs and process models.
Ebi can be used as a command-line utility, in ProM, or as a Rust library.

Ebi as a command-line interface tool

Download

Ready-to-use up-to-date binaries are available for Windows and Linux.

Ebi runs on Mac too; please see the manual on how to compile it in 5 easy steps.

Usage

Installation is not necessary: simply call the executable from a command line. Ebi does not use network access.
In Linux, you may need to give the executable execution permission.

An overview of the functions of Ebi and more information on its algorithms can be found in the manual.
Alternatively, you can execute Ebi with `--help' for an overview of commands.

Ebi in ProM

Ebi is also available in ProM. In the ProM Package Manager, install Ebi. Then, several Ebi commands can be used like any other ProM plug-in. As Ebi is extremely flexible in its inputs and outputs, please ensure you choose the correct input and output plug-in (there are many).

The manual contains an overview of Ebi commands that are available in ProM.

Ebi as a Rust library

Ebi is also available as a Rust library. More information can be found on crates.io.

Source

Ebi's source code is available from Github.

Commands

Ebi offers the following comands and techniques. Please refer to the manual for more information.

• Analyse all traces. Find all traces of a stochastic model.
• Analyse completeness. Estimate the completeness of an event log using species discovery.
• Analyse coverage. Find the most-likely traces that together cover a minimum probability.
• Analyse directly follows edge difference. The number of edges that differ between two directly follows graphs.
• Analyse medoid. Find the traces with the least distance to the other traces.
• Analyse minimum probability traces. Find all traces that have a given minimum probability.
• Analyse mode. Find a trace with the highest probability in a stochastic model.
• Analyse most likely traces. Find the traces with the highest probabilities.
• Analyse variety. Compute the variety of a stochastic language. That is, the average distance between two arbitrary traces in the language.
• Analyse non stochastic alignment. Compute alignments.
• Analyse non stochastic any traces. Compute whether the model has any traces.
• Analyse non stochastic bounded. Compute whether the model has a bounded state space.
• Analyse non stochastic cluster. Apply k-medoid clustering on a finite set of traces, without considering the stochastic perspective.
• Analyse non stochastic executions. Compute the executions of each transition of the model in the log.
• Analyse non stochastic infinitely many traces. Compute whether the model has infinitely many traces.
• Analyse non stochastic medoid. Find the traces with the least distance to the other traces, without considering the stochastic perspective.
• Association all trace attributes. Compute the association between the process and trace attributes.
• Association trace attribute. Compute the association between the process and a trace attribute.
• Conformance earth movers stochastic conformance. Compute Earth mover's stochastic conformance.
• Conformance earth movers stochastic conformance sample. Compute Earth mover's stochastic conformance with sampling.
• Conformance entropic relevance. Compute entropic relevance (uniform).
• Conformance jensen shannon. Compute Jensen-Shannon stochastic conformance.
• Conformance jensen shannon sample. Compute Jensen-Shannon stochastic conformance with sampling.
• Conformance unit earth movers stochastic conformance. Compute unit-earth movers' stochastic conformance.
• Convert finite stochastic language. Convert an object to a finite stochastic language.
• Convert labelled petri net. Convert an object to a labelled Petri net.
• Convert stochastic finite deterministic automaton. Convert an object to a stochastic deterministic finite automaton.
• Discover alignments. Give each transition a weight that matches the aligned occurrences of its label.
• Discover directly follows graph. Discover a directly follows graph.
• Discover occurrence labelled petri net. Give each transition a weight that matches the occurrences of its label; silent transitions get a weight of 1.
• Discover occurrence process tree. Give each leaf a weight that matches the occurrences of its label; silent leaves get a weight of 1.
• Discover uniform labelled petri net. Give each transition a weight of 1 in a labelled Petri net.
• Discover uniform process tree. Give each leaf a weight of 1 in a process tree.
• Discover non stochastic flower deterministic finite automaton. Discover a DFA that supports any trace with the activities of the log.
• Discover non stochastic flower process tree. Discover a process tree that supports any trace with the activities of the log.
• Discover non stochastic prefix tree deterministic finite automaton. Discover a DFA that is a prefix tree of the log.
• Discover non stochastic prefix tree process tree. Discover a process tree that is a prefix tree of the log.
• Information. Show information about a file.
• Probability explain trace. Compute the most likely explanation of a trace given the stochastic model.
• Probability log. Compute the probability that a stochastic model produces any trace of a log.
• Probability trace. Compute the probability of a trace in a stochastic model.
• Sample. Draw traces randomly from a model.
• Test bootstrap test. Test the hypothesis that the logs are derived from identical processes.
• Test log categorical attribute. Test the hypothesis that the sub-logs defined by the categorical attribute are derived from identical processes.
• Validate. Attempt to parse any file supported by Ebi. If you do not know the type the file should have, try `Ebi info`.
• Visualise graph. Visualise a file as a graph.
• Visualise text. Visualise a file as text.

Supported file formats

• Compressed event log (.xes.gz)
• Directly follows graph (.dfg)
• Deterministic finite automaton (.dfa)
• Directly follows model (.dfm)
• Stochastic directly follows model (.sdfm)
• Event log (.xes)
• Executions (.exs)
• Finite language (.lang)
• Finite stochastic language (.slang)
• Labelled petri net (.lpn)
• Language of alignments (.ali)
• Lo la petri net (.lola)
• Petri net markup language (.pnml)
• Stochastic deterministic finite automaton (.sdfa)
• Stochastic labelled petri net (.slpn)
• Process tree (.ptree)
• Portable document format (.pdf)
• Scalable vector graphics (.svg)
• Stochastic language of alignments (.sali)
• Stochastic process tree (.sptree)