On lazy and eager interactive reconfiguration

An interactive configuration tool needs to provide feedback to the user on possible further decisions while respecting constraints of the product being configured. In the presence of a large number of product features, it reduces the configuration effort if users can start from a default configuration and adapt only those features that are important to them. Hence, rather than completing an empty configuration (empty product), it is easier to move from one complete configuration to another (from one product to another). This paper shows how to provide tool support for this approach to interactive configuration. Two types of algorithms, based on recent advancements in SAT technology, are introduced: lazy and eager. While the eager provides more information to the user, the lazy scales to configuration models with tens of thousands of features. This is confirmed by an experimental evaluation carried out with the implemented prototype.

[1]  Barry O'Sullivan,et al.  Generating Corrective Explanations for Interactive Constraint Satisfaction , 2005, CP.

[2]  Inês Lynce,et al.  Towards efficient MUS extraction , 2012, AI Commun..

[3]  Mikolás Janota,et al.  Minimal Sets over Monotone Predicates in Boolean Formulae , 2013, CAV.

[4]  R. M. Jensen,et al.  FAST BACKTRACK-FREE PRODUCT CONFIGURATION USING A PRECOMPILED SOLUTION SPACE REPRESENTATION , 2004 .

[5]  Mikolás Janota,et al.  How to Complete an Interactive Configuration Process? , 2010, SOFSEM.

[6]  Krzysztof Czarnecki,et al.  SAT-based analysis of feature models is easy , 2009, SPLC.

[7]  Niklas Sörensson,et al.  An Extensible SAT-solver , 2003, SAT.

[8]  Krzysztof Czarnecki,et al.  The Variability Model of The Linux Kernel , 2010, VaMoS.

[9]  Krzysztof Czarnecki,et al.  Efficient synthesis of feature models , 2012, SPLC '12.

[10]  Pierre Marquis,et al.  Consistency restoration and explanations in dynamic CSPs Application to configuration , 2002, Artif. Intell..

[11]  Barry O'Sullivan,et al.  Finding Diverse and Similar Solutions in Constraint Programming , 2005, AAAI.

[12]  Douglas C. Schmidt,et al.  Automated diagnosis of feature model configurations , 2010, J. Syst. Softw..

[13]  Krzysztof Czarnecki,et al.  Feature Diagrams and Logics: There and Back Again , 2007, 11th International Software Product Line Conference (SPLC 2007).

[14]  Pierre-Yves Schobbens,et al.  Generic semantics of feature diagrams , 2007, Comput. Networks.

[15]  Mutsunori Banbara,et al.  A Compact and Efficient SAT-Encoding of Finite Domain CSP , 2011, SAT.

[16]  Kyo Chul Kang,et al.  Feature-Oriented Domain Analysis (FODA) Feasibility Study , 1990 .

[17]  Mikolás Janota,et al.  Proceedings of the Twenty-Third International Joint Conference on Artificial Intelligence On Computing Minimal Correction Subsets , 2022 .

[18]  Mikolás Janota,et al.  On Computing Minimal Equivalent Subformulas , 2012, CP.

[19]  Mikolás Janota.  Do SAT Solvers Make Good Configurators? , 2008, SPLC.

[20]  Hans Kleine Büning,et al.  Minimal Unsatisfiability and Autarkies , 2009, Handbook of Satisfiability.

[21]  Armin Biere,et al.  Managing SAT inconsistencies with HUMUS , 2012, VaMoS '12.