Adaptive Weighting of Structural Dependency and Textual Similarity in Software Architecture Recovery

Jae-Chul UM  Ki-Seong LEE  Chan-Gun LEE  

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E99-D   No.3   pp.756-759
Publication Date: 2016/03/01
Publicized: 2015/12/15
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2015EDL8160
Type of Manuscript: LETTER
Category: Software Engineering
Keyword: 
software architecture recovery,  module view,  dependency,  Newman modularity,  information entropy,  

Full Text: PDF>>
Buy this Article




Summary: 
Software architecture recovery techniques are often adopted to derive a module view of software from its source code in case software architecture documents are unavailable or outdated. The module view is one of the most important perspectives of software architecture. In this paper, we propose a novel approach to derive a module view by adaptively integrating structural dependency and textual similarity. Our approach utilizes Newman modularity and Shannon information entropy to determine the appropriate weights of the dependencies during the integration. We apply our approach to various open-source projects and show the experimental results validating the effectiveness of the approach.