Random forest-based research method for project stakeholder analysis

Abstract

The main purpose of this paper is the presentation of the Random Forest-based research method for project stakeholder analysis. This method can be a significant supplement to the solutions used so far in this field. Properly conducted stakeholder analyses are one of the most important conditions for the correct implementation of projects and achieving success, which may have different meanings and understanding for diverse individuals. It is necessary to thoroughly understand their needs, discover interrelationships and dependencies, as well as identify stakeholder who might display low or high interest in a given project. Therefore, complex and multifaceted research should be conducted on both the internal project environment and often complex issues related to the external environment. Existing methods, techniques and research tools are often insufficient, and it is worth paying attention to the possibilities of developing them to support the analyses more effectively. Thus, it is interesting to look for new research methods that could be used in the analysis of stakeholders and to run an analysis and assessment of their usefulness in evaluation studies on the ground of management sciences. The conducted study answers the following research question: Is there the need and possibility of using the method based on Random Forest in the analysis of project stakeholders? The proposed Random Forest-based research method enables conducting analyses using large collections of various data, obtaining reliable research results, relatively easy assessment of developed models and objective estimation of stakeholder interest misclassification and computational errors. This method is well suited for further development in order to meet new challenges of project stakeholder analysis. The paper presents not only theoretical analyses, but also the results of empirical studies. Important differences between the proposed method and other selected solutions in this field are synthetically pointed out. Finally, possible directions for further research in the field of stakeholder analysis methods development are indicated.