Power grid construction projects are proceeded on the basis of the multi-department collaboration. Besides, the whole process of project management involves a huge amount of data, which is difficult to disclose its inside correlation. Currently, the higher standards for power grid construction project management poses challenges on data management. Thus, it is necessary to further explore the inside data relationship in the process of project management, so as to address the data inaccuracy issue. In this paper, a correlation analysis method is proposed based on a modified apriori algorithm to study the correlation among the key indicators in the process of power grid construction project management. Specifically, the modified apriori algorithm by deploying dichotomizing search, can successfully decrease high computation demand while guaranteeing the analyzing accuracy. By establishing a data mining tool to analyze the indexes from the power grid construction projects in recent years, the pre-control and in-process control methods are proposed for the project management. In this regard, the paper offers a brand-new way to for the data management in the power grid construction projects.