Data-Driven Organizational Structure Optimization: Variable-Scale Clustering

With the continuous improvement of external data acquisition ability and computing power, data-driven optimization of organizational structure becomes an emerging technique for various enterprises to develop business performance and control management costs. This paper focuses on the management scale level discovery problem for the optimization of enterprise organizational structure. Firstly, according to the scale transformation theory, the scale level of the multi-scale dataset is defined. Then, a scale level discovery method based on the variable-scale clustering (SLD-VSC) is proposed. After determining management objectives, the SLD-VSC is able to recognize optimal management scale level and the scale characteristics of each management object clusters distributed in different management scale levels. The numerical experimental results illustrate that the proposed SLD-VSC is able to support enterprises improving their organizational structure by identifying the management scale levels from business data.