Research On The Configuration Analysis Of Financial Performance Of Medical Enterprises Based On Text Mining Technolog

Text mining is a natural language processing technique that explores textual information in depth.LDA topic model is a common approach to text mining with unsupervised feature.LDA topic model identifies potential topics in unordered text and mines the implicit information in the topics.fsQCA is a new approach to research in the field of management.The approach takes a configuration perspective that premised on the interplay between cause and effect,it enables the discovery of mechanisms of causal complexity in management practices and explores the nature of their operation.This paper explores the textual influences on the financial performance of healthcare companies using samples of 41 healthcare companies’ financial statement abstracts for the past 10 years,and proposes to discover the constructive relationships between the texts using a combination of LDA topic model and fsQCA.The study found that:(1)the LDA topic model was able to uncover four potential themes in the annual report summary,namely industry dynamics,financial position,equity information and business continuity;(2)the fsQCA approach resulted in nine groups of financial performance in terms of profitability,solvency and operating capacity,each with its own characteristics;(3)in terms of profitability,equity information was the main concern in measuring the profitability of healthcare companies;(4)in terms of solvency,business continuity was a better indicator of a healthcare company’s solvency;(5)in terms of operating capacity,the combination of financial position and business continuity is a stronger explanation of operating performance.Finally,the paper makes corresponding recommendations for different stakeholders.