Maintaining and analyzing production process definitions using a tree-based similarity measure

Autoren Reinhard Stumptner
Christian Lettner
Bernhard Freudenthaler
Josef Pichler
Wilhelm Kirchmayr
Ewald Traxler
Editoren E. Hüllermeier
M. Minor
Titel Maintaining and analyzing production process definitions using a tree-based similarity measure
Buchtitel Case-Based Reasoning Research and Development - Proc. ICCBR 2015
Typ Konferenzband
Verlag Springer
Serie Lecture Notes in Artificial Intelligence
Band 9343
ISBN 978-3-319-24585-0
DOI 10.1007/978-3-319-24586-7_25
Monat September
Jahr 2015
Seiten 366-380
SCCH ID# 1548
Abstract

In this work a Case-Based reasoning system for managing production processes, declarative production process definitions in particular, with main focus on analysis and maintenance is introduced whereby each process task is represented by a case. A single process task definition includes among other elements, formulas, represented by fragmental program code. To get a meaningful similarity function among such cases, a new fuzzy tree edit distance metric on the formulas’ abstract syntax tree has been developed. The fuzzy tree edit distance addresses two aspects of similarity – similarity in terms of similar structure and similarity in terms of similar wording. As such, the proposed method represents a multidisciplinary approach to production process maintenance that includes methods from Case-Based reasoning and code clone detection.