PDPlan: Uma Paralelização Eficiente de um Algoritmo de Mineração de Planos

Arlei L. da SilvaGuilherme P. PimentaWagner Meira Jr.

Plan mining is a key task for several business intelligence activities. In this paper we study the problem of mining large plan databases, in which states are measured by quality and actions have related costs. We aim to find, given an initial state, a sequence of actions that lead to the best final states, considering the compromise between cost and benefit of such actions and final states respectively. We employed a Filter-Stream parallelization model to treat huge plan databases in an efficient and scallable way, using Anthill.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: