Bringing Data Science to Practice: From Protype to Utilisation

Abstract

Data science and artificial intelligence have passed the stage of innovative trends. The applications in practice increase with every year with enterprises of all industry sectors creating new solutions utilising their data. However, there is much to learn for the enterprises, especially for those new to the implementation of information technology and data-based projects. Data science process models can assist in structuring such projects by giving ideal-typical project structures and assist with the provision of explanations, best practices, and concrete tools. One aspect which is rarely covered by data science process models is the utilisation of the results beyond their technical integration. This includes the risk of failing in operation due to missed requirements regarding affected employees or organisational aspects of the enterprises, especially their business processes. This paper provides an overview of relevant aspects for the integration of new data-based solutions into practice, i. e. the socio-technical system environment of the enterprise. Bridges to different project phases and results are shown to derive measures for integration. In addition, common tools for handling the arising challenges and tasks are listed and briefly discussed.