Loading...

Media is loading
 

Description

Data science projects can become very complex, due to the complexity of their content, but also due to the nature and composition of their stakeholders. There are several approaches to remedy this, e.g., canvases, which support ideation and common understanding. However, previous approaches are limited to single details or abstract too much, so that it is difficult to carry out entire projects successfully based on them. This paper describes one part of the design process, namely the derivation of the underlying ontology, of a new canvas that integrates both the overall project and detail steps. The ontology is mainly derived from CRISP-DM, literature review and project work.

Share

COinS
 
Jan 17th, 12:00 AM

Developing an ontology for data science projects to facilitate the design process of a canvas

Data science projects can become very complex, due to the complexity of their content, but also due to the nature and composition of their stakeholders. There are several approaches to remedy this, e.g., canvases, which support ideation and common understanding. However, previous approaches are limited to single details or abstract too much, so that it is difficult to carry out entire projects successfully based on them. This paper describes one part of the design process, namely the derivation of the underlying ontology, of a new canvas that integrates both the overall project and detail steps. The ontology is mainly derived from CRISP-DM, literature review and project work.