Discovery of fuzzy sequential patterns for fuzzy partitions in quantitative attributes

Abstract

In this paper we propose the Fuzzy Grids Based Sequential Patterns Mining Algorithm (FGBSPMA) to generate all fuzzy sequential patterns from relational database. In FGBSPMA, each quantitative attribute is viewed as a linguistic variable, and can be divided into many candidate 1-dim fuzzy grids. FGBSPMA is consisted of two phases: one is to generate all the large 1-fuzzy sequences, the other is to generate all the fuzzy sequential patterns. FGBSPMA is a efficiently fuzzy sequential patterns mining algorithm, because FGBSPMA scans database only once and applies proper operations on rows of tables to generate large fuzzy sequences and fuzzy sequential patterns. An example is given to illustrate a detailed process for mining the fuzzy sequential patterns from a specified relation. From this example, we can show efficiency and usefulness of FGBSPMA.