Data Mining system has a large number of data processing tasks. This paper mainly studies on task scheduling problem which takes the minimum total completion time based on the m equivalent processors and n tasks with priority. Firstly, we propose a data mining system multi-task scheduling model. Secondly, we adopt genetic algorithm solving this model. Due to initial seed may not be satisfied with the precedence constraint, this paper uses the inverse adjacency list to modify. Finally, we show this model and its algorithm are feasible and valid by the simulation experiment.