On Extracting Feature Models from Sets of Valid Feature Combinations.

Haslinger, Evelyn Nicole; Lopez-Herrejon, Roberto Erick; Egyed, Alexander

doi:10.1007/978-3-642-37057-1_5

On Extracting Feature Models from Sets of Valid Feature Combinations.

by Evelyn Nicole Haslinger, Roberto Erick Lopez-Herrejon, Alexander Egyed

Abstract:

Rather than developing individual systems, Software Product Line Engineering develops families of systems. The members of the software family are distinguished by the features they implement and Feature Models (FMs) are the de facto standard for dening which feature combinations are considered valid members. This paper presents an algorithm to automatically extract a feature model from a set of valid feature combinations, an essential development step when companies, for instance, decide to convert their existing product variations portfolio into a Software Product Line. We performed an evaluation on 168 publicly available feature models, with 9 to 38 features and up to 147456 feature combinations. From the generated feature combinations of each of these examples, we reverse engineered an equivalent feature model with a median performance in the low milliseconds.

Reference:

Evelyn Nicole Haslinger, Roberto Erick Lopez-Herrejon, Alexander Egyed, "On Extracting Feature Models from Sets of Valid Feature Combinations.", pp. 53-67, 2013.

Bibtex Entry:

@Conference{DBLP:conf/fase/HaslingerLE13,
  Title                    = {On Extracting Feature Models from Sets of Valid Feature Combinations.},
  Author                   = {Evelyn Nicole Haslinger and Roberto Erick Lopez-Herrejon and Alexander Egyed},
  Booktitle                = {16th International Conference on Fundamental Approaches to Software Engineering (FASE), Rome, Italy},
  Year                     = {2013},
  Pages                    = {53-67},

  Abstract                 = {Rather than developing individual systems, Software Product Line Engineering develops families of systems. The members of the software family are distinguished by the features they implement and Feature Models (FMs) are the de facto standard for dening which feature combinations are considered valid members. This paper presents an algorithm to automatically extract a feature model from a set of valid feature combinations, an essential development step when companies, for instance, decide to convert their existing product variations portfolio into a Software Product Line. We performed an evaluation on 168 publicly available feature models, with 9 to 38 features and up to 147456 feature combinations. From the generated feature combinations of each of these examples, we reverse engineered an equivalent feature model with a median performance in the low milliseconds.},
  Doi                      = {10.1007/978-3-642-37057-1_5},
  File                     = {On Extracting Feature Models from Sets of Valid Feature Combinations:Conferences\\FASE 2013 - On Extracting Feature Models from Sets of Valid Feature Combinations\\On Extracting Feature Models from Sets of Valid Feature Combinations.pdf:PDF},
  Keywords                 = {variability, FWF P21321-N15, FWF M1421-N15, EU IEF 254965}
}