A Real-Time Collaborative Testing Approach for Web Application: Via Multi-tasks Matching

Abstract

Crowdsourcing has emerged as a promising paradigm for processing tasks such as annotation, surveys and even testing through the manpower resources on the Internet. While the benefits of harnessing the collective efforts of individuals are obvious, less is known about the quality control of testing in a crowdsourcing environment, especially the existing crowdsourced collaborative testing that is formulated as an optimized problem and distribute tasks in a push-based way. The goal of this paper designs a real-time collaborative testing approach (PM2CT) consisting of a task partition algorithm and a greedy task assignment algorithm that dynamically selects task group from the test cases in each iteration, and assigns a test case (or task) to the appropriate tester by matching the difficulty of that task to the corresponding ability of the tester. Four experiments are designed and carried out in order to verify in terms of quality, efficiency, reliability and scalability.