A Natural Language Approach for Interacting with Software Architecture Knowledge

Software Architecture Knowledge Management (SAKM) has become a central research topic over the last years. This field of research aims to capture and use knowledge from the software architecture process to aid in future development processes and projects. If this knowledge is not stored to make it explicit, and not managed through knowledge management activities, it is prone to vaporization. The goal of this thesis is to explore current approaches to SAKM and to identify which shortcomings and possible solutions exist. Based on these findings, a new approach to SAKM employing natural language processing will be developed and demonstrated. To ensure the efficacy of the proposed approach, an evaluation of its capabilities will be conducted. The methodology employed for this thesis is the design science research methodology. The defined steps of this methodology will yield findings relevant to answering the three research questions.