Improving Enterprise Architecture Mining Methods

Enterprise architecture takes a step in the path of digital transformation by providing a comprehensive system with an integrated view of the organization. In traditional methods, architects manually prepare enterprise architecture products. The manual methods of producing enterprise architecture products have many challenges. One of the main challenges is the production of volumes of documentation, the complexity and difficulty of making enterprise architecture artifacts, and their costly and error-prone nature. In response to these challenges, enterprise architecture mining is proposed, automatically producing outputs and artifacts of enterprise architecture instead of manual modeling. Surveys show that enterprise architecture analysis still needs to be mature, and full automation is still far away. Some methods may be successful in some layers, but reaching a method that automatically provides an integrated enterprise architecture model with sufficient accuracy remains the challenge.This research will use model analysis methods to improve enterprise architecture mining. At first, the presented method extracts enterprise architecture models automatically. Detecting architecture smells reveals enterprise architecture debts in the next step. Ultimately, it improves the extracted models using refactoring methods. The proposed method can improve the accuracy and quality of the extracted models.