Software Construction Monitoring and Predicting for Human-Machine Pair Programming

Abstract

Pair programming is one of the promising techniques advocated in agile development paradigm, but it tends to be more costly than one person-based programming and to lack a rigorous principle for governing the cooperation of the two programmers. In this paper, we put forward a novel technique called Software Construction Monitoring and Predicting to study an intelligent and automatic approach to human-machine pair programming. Its aim is to automatically, dynamically monitor the process of software construction for fault detection and to predict the possible future contents of the software towards its error-free completion. We describe the theoretical foundation and frameworks for Software Construction Monitoring (SCM) and Software Construction Predicting (SCP), respectively. We also discuss how SCMP can support the Specification-Based programming paradigm. Finally, we use simple examples to illustrate how SCM and SCP can be supported.