From NLP to Taxonomy: identifying and classifying key functionality concepts of multi-level project planning and control systems

Analysis of literature and industry practices in applied planning and control systems reveals a notable lack of effective processes and stakeholders' understanding regarding the optimal use of these systems. These gaps underscore the urgent need for a refined understanding and discovery of the underlying concepts of existing systems to address the complex dynamics of the planning and control domain better. Therefore, this study employed a multi-step approach using advanced text-mining techniques and expert validation to address these issues. Sentence-Bidirectional Encoder Representations from Transformers (SBERT) for semantic analysis, hierarchical clustering, and word cloud visualization were applied to classify and validate project planning and
control system functionality concepts into coherent clusters. Furthermore, a robust taxonomy of functionality concepts was developed by meticulously analysing the findings as well as considering the domain experts' insights.
As a result, 148 project planning and control systems' functionalities were classified into 20 coherent clusters with an average 87% alignment rate. A robust taxonomy of these functionalities was then formulated, emphasizing their importance across various scheduling levels. This taxonomy captures the complexities of project planning and control systems, facilitating informed decision-making and the integration of diverse planning and control systems
to handle project complexities. The research significantly contributes to the field by clarifying the core concepts of project planning and control systems, making them more understandable and actionable for project stakeholders.