Information Technology Reference
In-Depth Information
A Discovery Method of Anteroposterior
Correlation for Big Data Era
Takafumi Nakanishi
Abstract.
In this paper, we present a new knowledge extraction method on Big data
era. We introduce new concepts, anteroposterior correlation, and propose an extrac-
tion method of anteroposterior correlation. The anteroposterior correlation means
the correlation based on the time anteroposterior relation. We consider that Het-
erogeneity, continuity, and visualization are the most critical features of Big data
analytics, which provides a scale and connection merits based on them. No current
data analysis methods are based on opened assumptions. Big data analytics provides
a new data analysis method based on opening assumptions. In this paper, we espe-
cially focus on an aspect of heterogeneity. We discover a correlation in considera-
tion of the continuity of time. By our method, we effectively discover relationships
between heterogeneous things, events and phenomena. The anteroposterior correla-
tions are represented in relative comparison with each conditional probability distri-
bution. The one of the features of our method is a measurement correlation by using
conditional probability. That is, we calculate the correlation relative by representing
all in conditional probability, no absolutely. Our method is determined higher corre-
lation by comparison to each heterogeneous thing, event and phenomenon. This is
the most important points on the Big data era. When you apply current association
rule extraction techniques, you obtain too big rule base to organize them. By our
method, we realize the one of the methods for decision mining.
1
Introduction
We information science researchers have already constructed data sensing, aggre-
gation, retrieval, analysis, and visualization environment by web portals, software,
Search WWH ::
Custom Search