The comprehensive energy system

which can coordinate multiple types of energy and be characterized by a deep integration of “cyber-physical-social” systems

is emerging. There is therefore an urgent need to conduct in-depth study on data science and big data mining for energy systems. This paper presents an initial discussion on data science and its applications in comprehensive energy systems. The fundamentals of data science

in particular the importance of the statistical learning theory and data quality

are discussed first. The new progresses in big data mining

such as deep learning

transfer learning and cross domain data fusion

are introduced then. Finally

a brief review is given on the applications of data mining techniques in energy systems; some research problems in energy system data mining

which require further attentions in future

are also discussed. This work is supported by National Basic Research Program of China(973 Program)(No. 2013CB228202)

National Natural Science Foundation of China(No. 51477151)and Specialized Research Fund for the Doctoral Program of Higher Education of China(No. 20120101110112).