Define architectures for which the model should be tailored and a structure for them

[achim-guldner]

In the paper, we defined the following reference/concrete architectures to address:

• Green AI procedure models
  • CRISP-DM
  • SCRUM
• Green AI life cycle models
  • MLOPS
  • TDSP
  • CDAC
• Green AI measurement models
  • power measurement methods
  • computation measurement methods
  • data measurement methods

We need to specify which architechtures we want to address and how to structure and classify them.

We also need to talk about where and how to include the criteria and indicators. Possibly in discussion (sect. 5) we could add a subsection about criteria and how to evaluate the impacts and then link to a criteria catalog that addresses each factor.

[achim-guldner]

We should also include the definitions and Link to the paper: https://dl.gi.de/server/api/core/bitstreams/3aa2d14a-c229-4887-9957-035486b4b6b6/content

maybe in the "why is this important" section of the introduction. And maybe also the picture . We could also add "tailoring" and links to tailored models there.