Abstract: A Data-Mining algorithm based on the Support-Confidence-Lift Framework is designed in this paper to mine strong association and correlation rules, which conform to the logic of Distribution Automation System, from the mass and poor-quality historical database of Distribution Automation System. And two of the four core Utility Indexes of Distribution Automation System combined with the historical database of a certain practical distribution grid are taken as the case to illustrate the application of this algorithm. Furthermore, the indexes calculated with or without the data-mining algorithm and their corresponding statistical data are compared and studied carefully. The comparison fully demonstrates that the powerful ability of the algorithm to excavate hidden information and relationships among mass data is valuable to the practical electrical distribution engineering.