2011 | OriginalPaper | Buchkapitel
A Practice Probability Frequent Pattern Mining Method over Transactional Uncertain Data Streams
verfasst von : Guoqiong Liao, Linqing Wu, Changxuan Wan, Naixue Xiong
Erschienen in: Ubiquitous Intelligence and Computing
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
In recent years, large amounts of uncertain data are emerged with the widespread employment of the new technologies, such as wireless sensor networks, RFID and privacy protection. According to the features of the uncertain data streams such as incomplete, full of noisy, non-uniform and mutable, this paper presents a probability frequent pattern tree called PFP-tree and a method called PFP-growth, to mine probability frequent patterns based on probability damped windows. The main characteristics of the suggested method include: (1) adopting time-based probability damped window model to enhance the accuracy of mined frequent patterns; (2) setting an item index table and a transaction index table to speed up retrieval on the PFP-tree; and (3) pruning the tree to remove the items that cannot become frequent patterns;. The experimental results demonstrate that PFP-growth method has better performance than the main existing schemes in terms of accuracy, processing time and storage space.