A first look at unfollowing behavior on GitHub

Many open source software projects rely on contributors to fix bugs and contribute new features. On GitHub, developers often broadcast their activities to followers, which may entice followers to be project contributors. It is important to understand unfollowing behavior, maintain current followers,...

Full description

Saved in:
Bibliographic Details
Main Authors: JIANG, Jing, LO, David, YANG, Yun, LI, Jianfeng, ZHANG, Li
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2019
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/4361
https://ink.library.smu.edu.sg/context/sis_research/article/5364/viewcontent/Unfollowing_Github_2019_afv.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
Description
Summary:Many open source software projects rely on contributors to fix bugs and contribute new features. On GitHub, developers often broadcast their activities to followers, which may entice followers to be project contributors. It is important to understand unfollowing behavior, maintain current followers, and attract some followers to become contributors in OSS projects.Our objective in this paper is to provide a comprehensive analysis of unfollowing behavior on GitHub. To the best of our knowledge, we present a first look at unfollowing behavior on GitHub. We collect a dataset containing 701,364 developers and their 4,602,440 following relationships in March 2016. We also crawl their following relationships in May 2013, August 2015 and November 2015. We conduct surveys, define potential impact factors, and analyze the correlation of factors with the likelihood of unfollowing behavior.Our main observations are: (1) Between May 2013 and August 2015, 19.8% of active developers ever unfollowed some users. (2) Developers are more likely to unfollow those who have fewer activities, lower programming language similarity, and asymmetric relationships. Our results give suggestions for developers to reduce the likelihood of being unfollowed by their followers, and attract researchers’ attention on relationship dissolution.