Systematic analysis of constellation-based techniques by using Natural Language Processing

Among the methodologies supporting new developments in sustainable problem identification and solution at the personal, organizational and social levels, the constellation techniques have gained increasing attention, offering tangible visualizations of system dynamics, learning in more detail about specific components, and thereby extracting valuable information knowledge in a specific topic. However, due to their complexity and interdisciplinary context, computational tools based on linguistics, computer science, and artificial intelligence, such as Natural Language Processing (NLP), perform a feasible approach to support data analysis and discussion. This study presents a systematic analysis and discusses the constellation-based techniques' principles, applications, potentialities, and limitations. An exhaustive investigation was performed to collect relevant information on digital databases, and NLP was applied to extract and process data, such as geographical location, classification of knowledge area, keywords extraction, and sentiment perception. Results indicate an upward trend in developing new studies related to constellation techniques within almost 20 countries mentioned in the publications. An extensive predominance of family constellations in psychology, medicine, and public health (94.17%) and Law and legal systems (97.96%) was observed. Also, it seems the publications are highly focused on problem-solving (74.58%). Nonetheless, it was observed an apparent inconsistency between principles and applications in the various constellation techniques. Moreover, the aspects of reproducibility remain insufficiently explored, as well as the ethical issues from facilitators regarding the coherence on how the constellation technique principles are applied. Furthermore, it is feasible to notice the growth of scientific interest in this approach and its benefits as a sustainable support tool for diagnosing problems and possible solutions, as well as, the implementation of social change processes, increasing the perception of systemic relationships in complex environments. The application of this technique in decision-making processes has the potential to be further exploited. To reach this goal, there is the need to progress integrative studies associated with the development of computational tools providing more objective, valuable, and sustainable results for a systemic analysis.