Based on the literatures of current research on carbon emission models in China and abroad, this paper uses the bibliometrics statistical method to investigate the data driven carbon emission research literatures published in CNKI and SCI from 2011 to 2020. In this research, we systematically analyzed the work done by predecessors from the aspects of industries, data models and data sets. This paper reveals the new research direction of carbon emission data models in future. We found a trend that the combination of different machine learning models and optimized traditional algorithms will be applied to the research on carbon emission model in future increasingly.