Visualización de la información por jerarquías

Abstract

At present analysis in databases is difficult due to various reasons: either because of large volumes stored, or internal structures in the data set. This causes the search for anomalies or situations of interest in a data set to be complex. Suppose in a product sales company this situation occurs: They want to know what level of classification (hierarchy) of products with high sales levels, say above 80% over the previous year. This consulting business called "trend in hierarchy" is locating those items in a hierarchy in the data submitted by a situation of interest to be analyzed. This means finding the relevant elements and the elements necessary to achieve the landmarks. This paper presents an analysis, design and implementation of a tool to solve the kind of business question posed. This consists in finding items of interest within the hierarchy tree of data cube dimension and then presents the results in visual representations recommended by information display scientists. There are 3 types of representations which are nodes maps, heat and Pastel Multi-Level maps. In addition to dashboards analyzing anomalies or points of interest in other levels of the hierarchy, which allows navigation on different granularity levels and other dimensions involved in the analysis