Research on Meteorological Data Quality Control Based on Association Rule Algorithm

Traditional data quality control algorithms can accurately assess the accuracy of changes in meteorological observation data. However, as they rely on boundary values derived from historical data for quality control, the sensitivity of these traditional algorithms is limited. To address these limitations, this paper proposes a novel association rule mining algorithm based on the interest model for meteorological data quality control. Our method not only identifies all rules with strong correlations but also outperforms non-Apriori class algorithms in terms of time performance, as demonstrated by real data validation. Subsequently, the association rule algorithm is utilized to mine historical meteorological data and extract all association item pairs to create a reference library. A meteorological data quality control model is then established based on this library. Through testing and verification, the study confirms that the quality control model based on the new association rule algorithm exhibits significantly enhanced detection rate, sensitivity, and performance, making it a highly feasible approach.