The risk assessment of debris flow is an effective method for disaster prevention and reduction. It is of great significance to carry out risk assessment of debris flow. This paper analyzes the characteristics of large-scale and multi-sourced big data of debris flow information, and the characteristics of space-time variability, uncertainty, randomness, and nonlinearity of debris flow systems. The selection of debris flow evaluation indexes, index weight determination and evaluation methods were analyzed, and the problems existing in the current debris flow evaluation were pointed out, and related solutions were discussed. It is pointed out that based on the Internet of Things and big data transmission, adopting multi-information fusion technology, using complex science system theory and artificial intelligence methods, it is the future trend to evaluate and forecast the debris flow"s time, space and scale with high precision and high timeliness.