Learning a process based on a set of example event traces

Develop a Java standalone application that adapts an implementation of a clustering algorithm to transitions between states and maps an obtained finite state representation onto a Little-JIL representation

"Automating Process Discovery through Event-Data Analysis" by Jonathan E. Cook and Alexander L. Wolf, ICSE 17, April 1995, Seattle, Washington, USA [pdf]

"Real-Time Training of Team Soccer Behaviors" by Keith Sullivan and Sean Luke [pdf]

Little-JIL papers and download [LJIL]

Develop a tool for model-driven design of Java implementations for control loops

Implement a tool that takes a Little-JIL definition in XML and combines a Java software system out of existing code snippets that corresponds to the definition

Develop a tool for simulation of Little-JIL process definitions using Kodkod

Implement a tool that takes a Little-JIL definition in XML, translates its constraints into Kodkod Java framework and obtains partial order plans by running the Kodkod solver

Kodkod papers and download [Kodkod]

Model JSD in Little-JIL

Use the Little-JIL plugin for Eclipse to create a runnable mock-up of JSD software process

Model Cleanroom in Little-JIL

Use the Little-JIL plugin for Eclipse to create a runnable mock-up of Cleanroom software process

Model OOP by Booch in Little-JIL

Use the Little-JIL plugin for Eclipse to create a runnable mock-up of OOP software process

Conduct a case study of ProMiner process mining tool by W.M.P. van der Aalst

Learn how to use the ProMiner tool, create mock-up event logs corresponding to a defined OOP Booch development process applied to a Java GUI calculator example and apply the ProMiner to the logs. Compare the mined Petri-Net model to the process you followed

Submission

Turn in the archive with the Eclipse project directory and a paper describing the design and evaluation electronically to the TRACS drop box.