Impact of Task Cycle Pattern on Project Success in Software Crowdsourcing

TL;DR

This paper empirically analyzes how different task cycle patterns in crowdsourcing affect project success, identifying four distinct lifecycle patterns that influence task completion and worker engagement.

Contribution

It introduces a novel empirical classification of task cycle patterns in crowdsourced software development based on lifecycle analysis.

Findings

Identified four distinct task cycle patterns: Prior, Current, Orbit, and Fresh.

Analyzed the impact of these patterns on project success metrics.

Provided insights into managing task workflows for better crowdsourcing outcomes.

Abstract

Crowdsourcing is becoming an accepted method of software development for different phases in the production lifecycle. Ideally, mass parallel production through Crowdsourcing could be an option for rapid acquisition in software engineering by leveraging infinite worker resource on the internet. It is important to understand the patterns and strategies of decomposing and uploading parallel tasks to maintain a stable worker supply as well as a satisfactory task completion rate. This research report is an empirical analysis of the available tasks' lifecycle patterns in crowdsourcing. Following the waterfall model in Crowdsourced Software Development (CSD), this research identified four patterns for the sequence of task arrival per project: 1) Prior Cycle, 2) Current Cycle, 3) Orbit Cycle, and 4) Fresh Cycle.