Asynchronous and Anticipatory Filter-Stream Based Parallel Algorithm for Frequent Itemset Mining

Abstract In this paper we propose a novel parallel algorithm for frequent itemset mining. The algorithm is based on the filter-stream programming model, in which the frequent itemset mining process is represented as a data flow controlled by a series producer and consumer components (filters), and the data flow (communication) between such filters is made via streams. When production rate matches consuption rate, and communication overhead between producer and consumer filters is minimized, a high degree of asynchrony is achieved. Our algorithm is built on this strategy − it employs an asynchronous candidate generation, and minimizes communication between filters by transfering only the necessary aggregated information. Another nice feature of our algorithm is a look forward approach which accelerates frequent itemset determination. Extensive experimental evaluation comproves the parallel performance and scalability of our algorithm.

• Type: Conference
• Year: 2004
• Where: PKDD
• Authors: Adriano Veloso, Wagner Meira Jr., Renato Ferreira, Dorgival Olavo Guedes Neto, Srinivasan Parthasarathy
• Keywords: Algorithm, Frequent Itemset, Frequent Itemset Mining, PKDD 2004