Exploring the Integration of Data Science Competencies in Modern Automation Frameworks: Insights for Workforce Empowerment

In today's professional world, a wide range of digital competencies are essential, with a focus on data science skills and knowledge of pre-processing data, developing features and building models. This expertise plays a crucial role in analyzing data sets of different sizes in various application contexts. Given the increasing demand for bringing these skills together in the work environment, which is characterized by an insufficient number of highly skilled individuals, smart solutions are being developed to strengthen these competencies. This study investigates the integration of data science competencies in the architecture of modern automation frameworks (AutoFM). In this way, it aims to encourage a discussion on the potential of these systems to facilitate the acquisition of data science skills by individuals, thus contributing to a broader discourse on workforce empowerment. For this purpose, a systematic literature search according to Webster and Watson was conducted to identify suitable automation frameworks corresponding to the data science competencies. Using the method of qualitative content analysis, the automation frameworks AutoPrep, AutoGluon-Tabular, AutoClust, and DeepEye were examined and interpreted to determine which of the functions mentioned could represent skills from the EDISON Data Science competence framework. The results show that the AutoFMs theoretically cover characteristics of the required skills of a data scientist such as data visualization in some competence groups, but cannot cover the entire competence. Some skills can therefore be performed by AutoFMs in an interaction between humans and technology, while other skills are not yet supported.