A Database for Mining Enterprise Architecture Models

Introduction

Since an Enterprise is a mostly pretty complex organization, it is hard for a single person to understand the structure and processes of the Enterprise. As a tool to visualize and model the Enterprise, Enterprise Architecture (EA) has been established. EA describes an Enterprise from an IT and business perspective and defines the structure and operation processes of the organization. EA has gained more and more relevance in the past years. Naturally a lot of different models for Enterprise Architecture have evolved.

Motivation

When someone decides to model an enterprise through EA, there are two options: The first option for this person is to find an existing EA model and adapt it to his Enterprise. The second option is to create a new EA model from scratch.

For researchers to find an existing and fitting model for their purpose can sometimes be hard. EA models developed by or on behalf of another company often remain (at least partly) unpublished. The published parts of the models are mostly not easily comprehensible and even harder to reconstruct for another enterprise.

This often leads researchers to the second option. They create a new Enterprise Architecture model. Thus there are plenty of different Enterprise Architecture models, in parts poorly documented, which makes it even harder for someone who is trying to find a fitting model.

People who work for a company and want to model their company through EA always have to create a new model. Nevertheless they would not have to create it from scratch, but could use similar EA models as an inspiration, if they wouldn’t be so unclearly documented.

The main motivation of this bachelor thesis is to make it easier for people to find the perfect EA model for every matter.

Implementation

In a first step I will perform a systematic literature review to assemble a collection of relevant papers. Since different EA models do not only use different modeling languages, but consider different layers of the organization and have different analysis concerns, certain models are completely unsuitable for some purposes. This is why in a second step I will categorize the models found in these papers regarding different aspects, e.g. used modeling language or considered layers of the Enterprise, and thus will derive a proper taxonomy for these models. In a third step I will construct a tool around an EA model database.

The tool should get requirements regarding characteristics of the desired model from the user, and return a suitable EA model from the database. This tool should ease the research and decision making process for everyone who is looking for a fitting Enterprise Architecture model. Subsequent the tool should be able to receive new EA models, categorize them in a semi-automatic way and add them to the database. In this way the database can easily be kept updated and thus provide people who look for one with a suitable state-of-the-art EA model.