Synthetic Minority Over-sampling TEchnique (SMOTE) for predicting software build outcomes
| aut.conference.type | Paper Published in Proceedings | |
| aut.researcher | Connor, Andrew Miles | |
| dc.contributor.author | Pears, R | |
| dc.contributor.author | Finlay, JA | |
| dc.contributor.author | Connor, AM | |
| dc.date.accessioned | 2014-04-21T06:42:48Z | |
| dc.date.available | 2014-04-21T06:42:48Z | |
| dc.date.copyright | 2014-07-01 | |
| dc.date.issued | 2014-07-01 | |
| dc.description.abstract | In this research we use a data stream approach to mining data and construct decision Tree models that predict software build outcomes in terms of software metrics that are derived from source code used in the software construction process. The rationale for using the data stream approach was to track the evolution of the prediction model over time as builds are incrementally constructed from previous versions either to remedy errors or to enhance functionality. As the volume of data available for mining from the software repository that we used was limited, we synthesized new data instances through the application of the SMOTE oversampling algorithm. The results indicate that a small number of the available metrics have significance for prediction software build outcomes. It is observed that classification accuracy steadily improves after approximately 900 instances of builds have been fed to the classifier. At the end of the data streaming process classification accuracies of 80% were achieved, though some bias arises due to the distribution of data across the two classes over time. | |
| dc.identifier.citation | Twenty-Sixth International Conference on Software Engineering and Knowledge Engineering (SEKE 2014) held at Hyatt Regency, Vancouver, Canada, 2014-07-01 to 2014-04-03 | |
| dc.identifier.issn | 2325-9000 | |
| dc.identifier.issn | 2325-9086 | |
| dc.identifier.uri | https://hdl.handle.net/10292/7131 | |
| dc.publisher | Knowledge Systems Institute Graduate School | |
| dc.relation.uri | http://www.ksi.edu/seke/Proceedings/seke/SEKE2014_Proceedings.pdf | |
| dc.rights | All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written consent of the publisher. | |
| dc.rights.accessrights | OpenAccess | |
| dc.subject | SMOTE | |
| dc.subject | Data stream mining | |
| dc.subject | Jazz | |
| dc.subject | Software metrics | |
| dc.subject | Software repositories | |
| dc.title | Synthetic Minority Over-sampling TEchnique (SMOTE) for predicting software build outcomes | |
| dc.type | Conference Contribution | |
| pubs.elements-id | 165699 | |
| pubs.organisational-data | /AUT | |
| pubs.organisational-data | /AUT/Design & Creative Technologies | |
| pubs.organisational-data | /AUT/Design & Creative Technologies/Interdisplinary Unit |