Open Access

ARTICLE

Evolution of Influential Developer’s Communities in OSS and its Impact on Quality

by Beenish Khan¹, Muhammad Rafiq Mufti², Asad Habib³, Humaira Afzal⁴, Mohammad Abdul Moiz Zia⁵, Afshan Almas⁶, Shahid Hussain^7,*, Bashir Ahmad¹

1 Department of Computer Science, COMSATS University, Islamabad, 45550, Pakistan
2 Department of Computer Science, COMSATS University, Vehari, 61100, Pakistan
3 Department of Computer Science, Kohat University of Science & Technology, Kohat, 26000, Pakistan
4 Department of Computer Science, Bahaddin Zakeria University, Multan, 60800, Pakistan
5 Department of Computer Science, University of Central Punjab, Lahore, 54590, Pakistan
6 Institute of Computer Science & Information Technology, The Women University Multan, 60800, Pakistan
7 Department of Computer and Information Science, University of Oregon, Eugene, 97401, Oregon, USA

* Corresponding Author: Shahid Hussain. Email:

(This article belongs to the Special Issue: Soft Computing Methods for Innovative Software Practices)

Intelligent Automation & Soft Computing 2021, 28(2), 337-352. https://doi.org/10.32604/iasc.2021.015034

Received 03 November 2020; Accepted 04 February 2021; Issue published 01 April 2021

Abstract

The high turnover of developers in the Open-Source Software (OSS) systems is due to the lack of restriction on a developer’s involvement and contributions. The primary developers start and administer an OSS project. However, they do not manage those who contribute. The literature shows that 80% of issues are resolved by 20% of developers when developing an OSS. Therefore, identifying influential developer communities is quite necessary for OSS stakeholders to reduce the efforts required to solve the issue through releases and predict quality. The purpose of this proposed empirical study is to explore influential communities by analyzing the relationship between their members as an OSS evolves and its impact on software quality. We performed several experiments with releases of three widely used OSS, namely “BIGDL,” “INCUBATOR-MXNET” and “RECOMMENDERS.” The major implications of the proposed study include; 1) The community development structure is not centralized and controlled, 2) Influential communities were observed in early releases of an OSS, 3) There is no guarantee of an influential community in the consecutive releases, 4) Notable developers are varied through the releases, and 5) The presence of influential communities in subsequent releases could lead to the maturity of an OSS.

---

Keywords

---