A Systematic Survey of Business Intelligence Literature Using Machine Learning Techniques

Abstract

Business intelligence (BI) comprises the strategies and technologies used by enterprises for the data analysis of business information for assisting its management and decision process. The principal aims of this study are: a) to complement the existing literature surveys in the BI area by identifying publications for the period 2007 to 2020, b) to classify these publications according to nine research strategies, c) to classify them according to various well-defined research topic categories, and d) apply machine learning techniques to assess their relevance with the BI field. We have collect 332 papers using ‘Google Scholar’ and a set of related keywords to the field of BI. The results show that most of the papers appeared during the years 2015 and 2017 [1]. The classifications of the literature based on research strategies and research topic indicate that most papers address formal theory and/or reviews and belong to the “benefits” research topic category. Finally, we attempt to increase the accuracy of the classification of aggregated information by utilizing machine learning techniques and estimating the “relevance” with respect to BI discipline. It appears that the overall best individual classifier is the SVM polynomial with ROSE sampling on 80% of the original data using the ‘TDM’ feature space, following by the Neural Networks with ROSE sampling on 80% of the original data using the ‘Topic’ feature space. We furthermore apply various ensemble methods, and we estimate and analyze their performance