The Weekly Mining of Fuzzy Patterns from Temporal Datasets

Abstract

The process of extracting fuzzy patterns from temporal datasets is a well known data mining problem. Weekly pattern is one such example where it reflects a pattern with some fuzzy time interval every week. This process involves two steps. Firstly, it finds frequent sets and secondly, it finds the association rules that occur in certain time intervals weekly. Most of the fuzzy patterns are concentrated as user defined. However, the probability of user not having prior knowledge of datasets being used in some applications is more. Thus, resulting in the loss of fuzziness related to the problem. The limitation of the natural language also bounds the user in specifying the same. This paper, proposes a method of extracting patterns that occur weekly in a particular fuzzy time frame and the fuzzy time frame is generated by the method itself. The efficacy of the method is backed by the experimental results