Shamim Ripon ¹, Musfiqur Rahman ², Javedul Ferdous ³, Md. Delwar Hossain ³

Affiliations
1 Depatment of Computer Science and Engineering, East West University, Dhaka, BD
2 Department of Computer Science and Software Engineering, Concordia University, Montreal, CA
3 Depatment of Computer Science and Engineering, East West University, Dhaka, IN

Abstract

Feature Tree represents all the features along with their relationship of a Software Product Line. Any defect in feature model can diminish the benefits of product line approach. Hence, the analysis of feature model plays a key role towards the success of any Software Product Line. This paper presents various analysis rules for cardinality-based feature model of both dead and false optional features. These rules are then verified by using Bayesian Network Based inference mechanism. Such verification not only confirms the analysis rules of the feature trees but also ensures the applicability of probabilistic information into the feature trees.

Keywords

Bayesian Network, Dead Feature, False Optional Feature, Feature Analysis, Software Product Line.

Full Text