The invention relates to a construction method for a big data power first-aid repair hotspot prediction system. The construction method comprises the following steps of 1, obtaining data of external systems and obtaining original data; 2, storing the original data in a database; 3, performing associative analysis by adopting different analysis methods according to different data types in the database; 4, performing method modeling according to an associative analysis result in the step 3 to form a hotspot regional prediction map; and 5, performing development on the hotspot regional prediction map to form an analysis display tool.