Expanding the intellectual capabilities of OLAP technology using neural networks

Abstract

The article highlights the main characteristics, features and structure of Online Analytical Processing systems based on the same technology that perform online analytical processing of data. This technology allows analysts to explore and navigate a multidimensional indicator structure called an online analytical processing cube (data cube). Indicators (measures) of data cube play an important role in the decision-making process. To solve certain problems, these measures often need to be classified or grouped. Moreover, empty measures are common in data cube. This fact negatively affects strategic decision making. Unfortunately, online analytical processing itself is not well suited for classifying, clustering, and predicting empty measures of data cube in the presence of large data. In this regard, today there is a need to use new technologies to solve such problems. Such technologies include neural networks. The article discusses the problem of integrating online analytical processing and a neural network, showing the possibilities and advantages of such integration. It mentions that in the case of big data, the integration of OLAP and neural networks is very effective in solving problems of classification, clustering and empty measure prediction of data cube. An architectural and technological model for the integration of online analytical processing and neural networks is presented.