Autor(es):
Cavique, Luís 
Data: 2007
Identificador Persistente: http://hdl.handle.net/10400.2/1954
Origem: Repositório Aberto da Universidade Aberta
Assunto(s): Data mining; Frequent sequence mining; Graph theory
Detalhes do Documento
A network algorithm to discover sequential patterns
Descrição
This paper addresses the discovery of sequential patterns in very large databases. Most of the existing algorithms use lattice structures in the space search that are very demanding computationally. The output of these algorithms generates a large number of rules. The aim of this work is to create a swift algorithm for the discovery of sequential patterns with a low time complexity. In this work, we also want to define tools that allow us to simplify the work of the final user, by offering a new visualization of the sequences, while bypassing the analysis of thousands of association rules.
This paper addresses the discovery of sequential patterns in very large databases. Most of the existing algorithms use lattice structures in the space search that are very demanding computationally. The output of these algorithms generates a large number of rules. The aim of this work is to create a swift algorithm for the discovery of sequential patterns with a low time complexity. In this work, we also want to define tools that allow us to simplify the work of the final user, by offering a new visualization of the sequences, while bypassing the analysis of thousands of association rules.
Tipo de Documento
Parte ou capítulo de livro
Idioma Inglês
Idioma Inglês
| Financiadores do RCAAP | |||||||
|
|
|
|
|
|
||