Data acquisition in Biology and Astronomy has seen unprecedented growth in volume since the turn of the century. It will not be an exaggeration to state that the needs of these two sciences are pushing computer science research to new frontiers. The focus of this paper is astronomy, which since inception of Virtual Observatory and commissioning of massive sky surveys is gasping for knowledge in data deluge. Astrocomputing, which subsumes Astroinformatics, is a recent multi-disciplinary field of research with computer science and astronomy at the core. In this article we dwell upon the opportunities and challenges for machine learning and data mining research thrown open by this emerging discipline. We present a case study of an ongoing work on exploratory analysis of unclassified light curves. Though scientific analysis and interpretation of the results of the study are pending, the exercise demonstrates the merit of customized exploratory approach for study. The approach is general and can be applied to light curves obtained from any survey. Owing to the gargantuan scale of astronomy data processing requirements, we discuss scalability of the proposed method.