On improving green mining for energy-aware software analysis.

Consumer demand for longer lasting battery life in mobile computers, as well as industry interest in energy efficient cloud infrastructure, creates a need for hardware and software energy efficiency improvements. One way to tackle this problem is from a software perspective. If it were known which software changes influenced energy consumption, then tools could be created to help software professionals create more energy efficient software. The process of extracting energy consumption information, Green Mining, is time demanding because researchers must run many tests, with sufficient coverage, on each revision in a software product multiple times. The time required for testing acts as a barrier to extracting energy consumption measurements from new software systems. Therefore, this work proposes, implements, and evaluates a search-based approximation method that trades some precision for a speed-up in the mining process. This speed-up enables researchers to study additional software systems that were too costly to investigate before.