STUDY OF DIFFERENT DATA SCIENCE APPROACHES

Abstract

For the purpose of developing new algorithms, improving data models, and producing sophisticated analytics, data science has made large research investments. However, authors have not frequently addressed the organisational and socio-technical difficulties that arise when carrying out a data science project. These difficulties include the absence of a defined vision and goals, the overemphasis on technical issues, the inadequacy of ad hoc projects, and the uncertainty of responsibilities in data science. There haven't been many methods proposed in the literature to deal with this kind of issue; some of them go as far back as the middle of the 1990s, so they aren't up to speed with the most recent developments in big data and machine learning technology. However, fewer approaches offer a complete framework. We'll discuss the necessity to develop a more thorough technique for working on data science projects in this piece. We first research ways that have been written about in the literature and group them into four categories based on their focuses: project, team, data, and information management. Last but not least, we offer a conceptual framework that describes the essential characteristics that a methodology for managing data science activities from a broad viewpoint should have. This framework could serve as a guide for other academics as they develop new data science methods or update existing ones