Evolution and scientific visualization of Machine learning field

Abstract

This article provides a retrospective and understanding of the development of automatic learning methods. The beginnings are visualized as a discipline within Computer Sciences in the subcategory of Artificial Intelligence, its development and the current transfer of knowledge to other areas of Engineering and its industrial applications. Based on the publications about machine learning and its application contained in the Web of Science database, records from 1986 to 2017 are downloaded. After a description of the technological profile, a new approach is introduced to the classification of a discipline based on the year of appearance of those terms that define it. Mining of technological texts and network theory has been applied to extract the terms and interpret their evolution. They are the those that define the stages of emergence, development and maturation of the discipline Machine learning. The novelty of this approach lies in the technical nature of applied research in Machine Learning, which aims to be a guide for the development of future engineering applications and to make technology transfer toindustry visible.