Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/76569
Title: StiCProb : a novel feature mining approach using conditional probability
Authors: Tang, YT 
Leung, H 
Keywords: Software product line
Variability
Feature mining
Program slicing
Issue Date: 2017
Publisher: Institute of Electrical and Electronics Engineers
Source: 24th IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER), Klagenfurt, Austria, Feb 20-24, 2017, p. 45-55 How to cite?
Abstract: Software Product Line Engineering is a key approach to construct applications with systematical reuse of architecture, documents and other relevant components. To migrate legacy software into a product line system, it is essential to identify the code segments that should be constructed as features from the source base. However, this could be an error-prone and complicated task, as it involves exploring a complex structure and extracting the relations between different components within a system. And normally, representing structural information of a program in a mathematical way should be a promising direction to investigate. We improve this situation by proposing a probability-based approach named StiCProb to capture source code fragments for feature concerned, which inherently provides a conditional probability to describe the closeness between two programming elements. In the case study, we conduct feature mining on several legacy systems, to compare our approach with other related approaches. As demonstrated in our experiment, our approach could support developers to locate features within legacy successfully with a better performance of 83% for precision and 41% for recall.
URI: http://hdl.handle.net/10397/76569
ISBN: 978-1-5090-5501-2
Appears in Collections:Conference Paper

Access
View full-text via PolyU eLinks SFX Query
Show full item record

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.