A metric description generator for Enterprise Architecture models

The discipline of Enterprise Architecture (EA) is an established approach to model and manage the interaction of businesses processes and IT in an organization. Thereby, the EA model as a central artifact of EA is subject to a continuous evolution caused by multiple sources of changes. This requires a lot of effort in controlling and managing the evolution of the EA model, which leads to very time-consuming and error-pron tasks. Additionally, the need to measure certain qualities of the EA model are of great importance for both the evolution of the EA model and for the enterprise. One option to measure the EA model’s quality is the calculation of EA maintenance KPI’s. For this purpose the SWC chair designed a DSL to describe the measurement of KPI’s in an EA model.

The DSL is split into two major parts. First, it describes which components of an EA model are queried and how they are calculated with each other to measure the KPI. Second, the DSL describes how the measured value will be interpreted, i.e. on which scale the value will be interpreted as well as the thresholds for a good, average or bad value of the KPI. The outcome of these KPI calculations are used in a continuous delivery pipeline for EA models. For example, if a certain KPI is calculated poorly, the new EA model is not allowed to be deployed as the organization wide EA model.

The DSL is designed in yaml, respectively json format. Therefore, it is a difficult task to create a new EA metric description in the DSL. Manually creating a description is error prone due to typos, missing or incorrect key words, etc. For this, a user-friendly generator of such metric descriptions will be of great value. The conceptualization and realization of such a generator will be task of this thesis.

This thesis is available as both bachelor or, with some extensions, as a master thesis.

 

References:

  • A. Sabau, S. Hacks, A. Steffens: Implementation of a Continuous Delivery Pipeline for EA Model Evolution, (in Review). Feb. 2020
  • Type: Bachelor Thesis, Master Thesis
  • Status: Open

Supervisor