Re-mining method of painted ceramic sculpture artistic material based on association rule algorithm

Abstract

This paper brings out an improved approach for re-mining colored ceramic sculpture art materials employing an improved association rule algorithm. The approach optimizes the Apriori algorithm with a genetic algorithm integration and a triangular matrix for data storage, which decreases the algorithm's runtime to Address the vast search spaces and susceptibility to local optima challenges. The primary improvements comprise a pre-pruning strategy that filters irrelevant data early streamlines the search procedure and enhances computational efficiency. Experimental outputs depict that the optimized algorithm generated over 5000 frequent itemsets in 25 s, while the traditional Apriori algorithm only managed to produce 1000 itemsets in 30 s. Furthermore, the confidence levels for association rules mined by the enhanced algorithm remained consistently above 0.7, even as the number of rules reached 250, whereas the confidence of the traditional algorithm declined to around 3.5 for the same rule count. These reports highlight the optimized algorithm's ability to achieve high efficiency and predictive accuracy and provide reliable data mining outputs in structured datasets and complex network environments. This method holds significant potential for supporting educators, researchers, and students by enabling efficient and precise extraction of diverse artistic materials, facilitating a deeper understanding and appreciation of colored ceramic sculpture art.