Using multivariate time series and association rules to detect logical change coupling

An empirical study

Gerardo Canfora, Michele Ceccarelli, Luigi Cerulo, Massimiliano Di Penta

Research output: Chapter in Book/Report/Conference proceedingConference contribution

38 Citations (Scopus)

Abstract

In recent years, techniques based on association rules discovery have been extensively used to determine changecoupling relations between artifacts that often changed together. Although association rules worked well in many cases, they fail to capture logical coupling relations between artifacts modified in subsequent change sets. To overcome such a limitation, we propose the use of multivariate time series analysis and forecasting, and in particular the use of Granger causality test, to determine whether a change occurred on a software artifact was consequentially related to changes occurred on some other artifacts. Results of an empirical study performed on four Java and C open source systems show that Granger causality test is able to provide a set of change couplings complementary to association rules, and a hybrid recommender built combining recommendations from association rules and Granger causality is able to achieve a higher recall than the two single techniques.

Original languageEnglish
Title of host publicationIEEE International Conference on Software Maintenance, ICSM
DOIs
Publication statusPublished - 20 Dec 2010
Externally publishedYes
Event2010 IEEE International Conference on Software Maintenance, ICSM 2010 - Timisoara
Duration: 12 Sep 201018 Sep 2010

Other

Other2010 IEEE International Conference on Software Maintenance, ICSM 2010
CityTimisoara
Period12/9/1018/9/10

Fingerprint

Association rules
Time series
Time series analysis

Keywords

  • Change coupling
  • Empirical study
  • Impact analysis
  • Multivariate time series
  • Software evolution

ASJC Scopus subject areas

  • Software

Cite this

Canfora, G., Ceccarelli, M., Cerulo, L., & Di Penta, M. (2010). Using multivariate time series and association rules to detect logical change coupling: An empirical study. In IEEE International Conference on Software Maintenance, ICSM [5609732] https://doi.org/10.1109/ICSM.2010.5609732

Using multivariate time series and association rules to detect logical change coupling : An empirical study. / Canfora, Gerardo; Ceccarelli, Michele; Cerulo, Luigi; Di Penta, Massimiliano.

IEEE International Conference on Software Maintenance, ICSM. 2010. 5609732.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Canfora, G, Ceccarelli, M, Cerulo, L & Di Penta, M 2010, Using multivariate time series and association rules to detect logical change coupling: An empirical study. in IEEE International Conference on Software Maintenance, ICSM., 5609732, 2010 IEEE International Conference on Software Maintenance, ICSM 2010, Timisoara, 12/9/10. https://doi.org/10.1109/ICSM.2010.5609732
Canfora G, Ceccarelli M, Cerulo L, Di Penta M. Using multivariate time series and association rules to detect logical change coupling: An empirical study. In IEEE International Conference on Software Maintenance, ICSM. 2010. 5609732 https://doi.org/10.1109/ICSM.2010.5609732
Canfora, Gerardo ; Ceccarelli, Michele ; Cerulo, Luigi ; Di Penta, Massimiliano. / Using multivariate time series and association rules to detect logical change coupling : An empirical study. IEEE International Conference on Software Maintenance, ICSM. 2010.
@inproceedings{34726ba21a8b40e5becdcdb93da4514a,
title = "Using multivariate time series and association rules to detect logical change coupling: An empirical study",
abstract = "In recent years, techniques based on association rules discovery have been extensively used to determine changecoupling relations between artifacts that often changed together. Although association rules worked well in many cases, they fail to capture logical coupling relations between artifacts modified in subsequent change sets. To overcome such a limitation, we propose the use of multivariate time series analysis and forecasting, and in particular the use of Granger causality test, to determine whether a change occurred on a software artifact was consequentially related to changes occurred on some other artifacts. Results of an empirical study performed on four Java and C open source systems show that Granger causality test is able to provide a set of change couplings complementary to association rules, and a hybrid recommender built combining recommendations from association rules and Granger causality is able to achieve a higher recall than the two single techniques.",
keywords = "Change coupling, Empirical study, Impact analysis, Multivariate time series, Software evolution",
author = "Gerardo Canfora and Michele Ceccarelli and Luigi Cerulo and {Di Penta}, Massimiliano",
year = "2010",
month = "12",
day = "20",
doi = "10.1109/ICSM.2010.5609732",
language = "English",
isbn = "9781424486298",
booktitle = "IEEE International Conference on Software Maintenance, ICSM",

}

TY - GEN

T1 - Using multivariate time series and association rules to detect logical change coupling

T2 - An empirical study

AU - Canfora, Gerardo

AU - Ceccarelli, Michele

AU - Cerulo, Luigi

AU - Di Penta, Massimiliano

PY - 2010/12/20

Y1 - 2010/12/20

N2 - In recent years, techniques based on association rules discovery have been extensively used to determine changecoupling relations between artifacts that often changed together. Although association rules worked well in many cases, they fail to capture logical coupling relations between artifacts modified in subsequent change sets. To overcome such a limitation, we propose the use of multivariate time series analysis and forecasting, and in particular the use of Granger causality test, to determine whether a change occurred on a software artifact was consequentially related to changes occurred on some other artifacts. Results of an empirical study performed on four Java and C open source systems show that Granger causality test is able to provide a set of change couplings complementary to association rules, and a hybrid recommender built combining recommendations from association rules and Granger causality is able to achieve a higher recall than the two single techniques.

AB - In recent years, techniques based on association rules discovery have been extensively used to determine changecoupling relations between artifacts that often changed together. Although association rules worked well in many cases, they fail to capture logical coupling relations between artifacts modified in subsequent change sets. To overcome such a limitation, we propose the use of multivariate time series analysis and forecasting, and in particular the use of Granger causality test, to determine whether a change occurred on a software artifact was consequentially related to changes occurred on some other artifacts. Results of an empirical study performed on four Java and C open source systems show that Granger causality test is able to provide a set of change couplings complementary to association rules, and a hybrid recommender built combining recommendations from association rules and Granger causality is able to achieve a higher recall than the two single techniques.

KW - Change coupling

KW - Empirical study

KW - Impact analysis

KW - Multivariate time series

KW - Software evolution

UR - http://www.scopus.com/inward/record.url?scp=78650109975&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=78650109975&partnerID=8YFLogxK

U2 - 10.1109/ICSM.2010.5609732

DO - 10.1109/ICSM.2010.5609732

M3 - Conference contribution

SN - 9781424486298

BT - IEEE International Conference on Software Maintenance, ICSM

ER -