Design data model

From EERAdata Wiki
Jump to: navigation, search

A data model organizes data in an abstract way and describes how the data relate to real world entities. In a research context, data can be values measured in an experiment. The data model describes how the data are related to each other. Often, certain data can be combined into a single entity with some attributes, e.g. one observation with several values that are observed. The page Understanding Simple Data Models: A Comprehensive Guide provides an introduction to data modelling. An important tool in data modelling is a graphical represenation of the different data and their relationships. A generalized modelling language coming with a graphical representation is the Unified Modelling Language.

We illustrate the data model with a specific example taken from use case 2. In this use case, an inventory of energy communities and their projects is FAIRified. Real world objects in that context are the communities coming with information such as street addresses and foundation dates and projects, which are characterized by capacities, locations, dates of commissioning etc. In the abstract data model, the data is organzined into two containers or classes. Both classes are linked to each other with a relationship stating that project A is owned by community B. Additional classes can be introduced with the aim to answer specific research questions. The figure shows a UML representation. Shown on the left hand is the class for initiatives which include attributes such as the name of the initiative, the address and geo-location, the website, its year of foundation and dissolution, etc. On the right hand side, the class of production units includes attributes such as location of the unit, type of technology and energy product used or provided, capacity, and equipment specifications. Both classes are connected by a ownership class, which specifies which initiative owns which production unit.