Unveiling Entrepreneurship in Chinese State-Owned Enterprises: A Computational Linguistic Analysis.

Entrepreneurship constitutes a pivotal facet in business practice and research, and state-owned enterprises (SOEs) stand as integral components within the Chinese economy. Diverging from private enterprises, Chinese SOEs, instead of being spearheaded by their entrepreneurs, find their origin in the state apparatus and fall under governmental purview. The board of directors, in turn, is typically led and appointed by the Communist Party of China as well. Given this backdrop, delving into the behavioral dynamics of entrepreneurship and its consequential impact on firm performance emerges as a compelling and yet unresolved inquiry. However, the scarcity of overt personal attributes impedes a direct assessment of the personalities and entrepreneurial traits inherent in these leaders. Therefore, we draw upon 80 public speeches delivered by leaders of central SEOs, employing computational linguistic methods to formulate two distinct text analysis models: the LDA topic model and the Word2Vec language model. Through these models, a quantitative analysis framework for entrepreneurship is established, elucidating the nuanced dimensions of entrepreneurship in Chinese SOEs. The research delineates 5 general and 13 detailed perspectives on entrepreneurship, measured through corresponding indicators. Utilizing regression analysis on the presented entrepreneurship traits and collated enterprise indicators, our measure of entrepreneurship traits establishes a robust foundation for future theoretical advancements concerning the impact of entrepreneurial personality traits. The specific findings contribute to an enhanced comprehension of how entrepreneurship influences the profitability and accountability of enterprises. [ABSTRACT FROM AUTHOR]