关键词: 高级模式挖掘, 最大模糊模式, 模糊支持度, 核心-牵引模式结构, 模糊修剪策略

Abstract: Abstract: The efficient mining for potentially useful but hidden information from large datasets and the representation of the embedded information by using proper structure are important in advanced pattern mining. Combinatorial explosion and the effectiveness of mining results are the essential challenges of meaningful pattern extraction. The existing algorithms cannot entirely solve these issues is because that a huge amount of candidate subsets have been generated and the items have the features of uncertainty. In this paper, we have proposed core-(second-order-effect) pattern structure, FFP-Tree structure, and fuzzy constraints to ensure the mining efficient of extracted patterns. The proposed maximal FFPs mining algorithm only need to scan the dataset only once and it also can reduce pattern generation based on fuzzy pruning strategy. The experimental results gained from the benchmark datasets analysis reveal that the proposed maximal FFPs algorithms has outstanding performance by comparing with PADS and FPMax* algorithms.

Key words: Advanced pattern mining, Maximum fuzzy pattern, Fuzzy support, core-(second-order-effect) pattern structure, Fuzzy pruning strategy