Development of Efficient and Robust Linkage Pattern Mining for Multiple Sequential Data

作     者：Maeda, Kyosuke Yokota, Issei Okada, Yoshifurni Lee, Saerom 

作者机构：Division of Science for Creative Emergence Kagawa University Graduate School 2-1 Saiwai-cho Kagawa Takamatsu760-8523 Japan Division of Science for Creative Emergence Kagawa University Graduate School 2-1 Saiwai-cho Kagawa Takamatsu760-8523 Japan College of Information and Systems Muroran Institute of Technology 27-1 Mizumoto-cho Hokkaido Muroran050-8585 Japan Faculty of Engineering and Design Kagawa University 1-1 Saiwai-cho Kagawa Takamatsu760-8523 Japan 

出 版 物：《IAENG International Journal of Computer Science》 (IAENG Int. J. Comput. Sci.)

年 卷 期：2025年第52卷第1期

页      面：223-232页

核心收录：

主　　题：closed itemset EMMA interval graph linkage pattern sequential pattern mining 

摘      要：Linkage pattern mining is a method used to extract frequently occurring patterns from multiple sequential data without considering the similarity or correlation between frequent patterns. Therefore, it is expected to be a promising approach for disease prediction and voice data analysis. In the previous method, closed itemset mining was introduced in the linkage pattern-mining algorithm to ensure robustness against noise. Although this method can extract linkage patterns from noisy artificial dataseis, a reduction in computation time and stringent parameter settings are essential for its practical application to real data. In this study, we employed Episode Mining using Memory Anchor algorithm for frequent pattern mining to overcome the limitations of the previous method. The objective of this study is to develop a new robust linkage pattern-mining method that is more applicable to real data, A performance comparison between the previous and proposed methods using artificial datasets showed that the proposed method achieved a reduction in computation time while maintaining an extraction accuracy that is comparable to that of the previous method, particularly on noisy artificial datasets. © (2025), (International Association of Engineers). All rights reserved.