Toward Dynamic Model Association through Semantic Analytics: Approach and Evaluation

Abstract

Business Architecture (BA) is often used as a blueprint that aligns an enterprise's capabilities and processes with its strategic objectives and structures. Conventionally, the alignment is established manually by subject matter experts through the association between model elements based on their domain knowledge; while the quality of the alignment is high, the effort is costly and time-consuming. To overcome these issues, we propose a novel yet complementary approach wherein associations between model elements are automatically and dynamically established and maintained through semantic analytics, such as natural language understanding and synonym tag generation. In this paper, we describe how the dynamic model association is implemented in IBM's CBM.next project, evaluate initial results from the deployment, and discuss directions for future research.