This paper describes a method for estimating the salience of features comprising conceptual descriptions of software artifacts. Salience estimates are used in a model analyzing the similarity between such descriptions so as to promote the analogical reuse of the artifacts described by them. Salience is conceived as belief on the dominance of a feature, which is defined on the basis of three general properties that a feature may have in a conceptual model, namely the abstractness, the characteristicity and the causality. This belief is measured according to evidence inherent in conceptual schemas organizing software repositories.
CITATION STYLE
Spanoudakis, G., & Constantopoulos, P. (1994). On evidential feature salience. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 856 LNCS, pp. 153–162). Springer Verlag. https://doi.org/10.1007/3-540-58435-8_180
Mendeley helps you to discover research relevant for your work.