Open AccessFuzzy analysis and prediction of commit activity in open source software projects
Author(s) -
Saini Munish,
Kaur Kuljit
Publication year - 2016
Publication title -
iet software
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.305
H-Index - 43
eISSN - 1751-8814
pISSN - 1751-8806
DOI - 10.1049/iet-sen.2015.0087
Subject(s) - commit , autoregressive integrated moving average , computer science , software , time series , fuzzy logic , correctness , set (abstract data type) , fuzzy set , predictive modelling , data mining , machine learning , artificial intelligence , database , operating system , algorithm , programming language
Autoregressive integrated moving average (ARIMA) models are the most commonly used prediction models in the previous studies on software evolution prediction. This study explores a prediction method based on fuzzy time series for predicting the future commit activity in open source software (OSS) projects. The idea to choose fuzzy time series based prediction method is due to the stochastic nature of the OSS development process. Commit activity of OSS project indicates the activeness of its development community. An active development community is a strong contributor to the success of OSS project. Therefore commit activity prediction is an important indicator to the project managers, developers, and users regarding the evolutionary prospects of the project in future. The fuzzy time series‐based prediction method is of order three and uses time variant difference parameter on the current state to forecast the next state data. The performance and suitability of computational method are examined in comparison with that of ARIMA model on a data set of seven OSS systems. It is found that the predicted results of the computational method outperform various ARIMA models. Towards the end, a commit prediction model is used for each project to analyse the trends in their commit activity.