Dimension Debasing towards Minimal Search Space Utilization for Mining Patterns in Big Data

Abstract

Data mining algorithms generally produce patterns which are interesting. Such patterns can be used by domain experts in order to produce business intelligence. However, most of the existing algoritms that can not properly work for uncertain data. Keeping uncertain data’s characteristics in mind, it can be said that they do have more search space with existing algorithms. In this paper we proposed a method that can be used to reduce search space besides helping in producing patterns from uncertain data. The proposed method is based on MapReduce programming framework that works in distributed environment. The method essentially works on big data which is characterized by velocity, volume and variety. The proposed method also helps users to have constraints so as to produce high quality patterns. Such patterns can help in making well informed decisions. We built a prototype application that demonstrates the proof of concept. The empirical results are encouraging in mining uncertain big data in the presence of constraints.
DOI: 10.17762/ijritcc2321-8169.150808