### Abstract

In this paper, we propose an association rule A ) Correl(X, Y ) that handles a correlation function, where A is the prerequisite and Correl(X, Y ) is the correlation between variables X and Y . With this extension, we can find conditions whose correlation of arbitrary two variables is high (or low) from a given data set. Furthermore, in order to distinguish statistically significant correlation, we define the rule A ) TestCorrel(X, Y ) which holds the result of the correlation significance test in the conclusion section, where TestCorrel(X, Y ) is a p-value of no-correlation test between X and Y . In order to confirm the feasibility of the proposed method, a case study using software development data was conducted. We found that it is possible to distinguish projects that are suitable for predicting development effort and those that are not.

Original language | English |
---|---|

Pages (from-to) | 47-53 |

Number of pages | 7 |

Journal | Computer Software |

Volume | 36 |

Issue number | 3 |

DOIs | |

Publication status | Published - Jan 1 2019 |

### Fingerprint

### ASJC Scopus subject areas

- Software

### Cite this

*Computer Software*,

*36*(3), 47-53. https://doi.org/10.11309/jssst.36.3_47