Research on measurement and disequilibrium of manufacturing digital transformation: Based on the text mining data of A-share listed companies
Chong Li,Guoqiong Long,Shuai Li
DOI: https://doi.org/10.3934/dsfe.2023003
2023-01-01
Data Science in Finance and Economics
Abstract:Quantitative analysis of digital transformation is an important part of relevant research in the digital field. Based on the annual report data of China's manufacturing listed companies from 2011 to 2019, this study applies cloud computing to the mining and analysis of text data, and uses the Term Frequency-Inverse Document Frequency method under machine learning to measure the digital transformation index value of manufacturing enterprises. The results show that: (1) On the whole, the current pace of digital transformation of manufacturing enterprises continues to accelerate, and the digital transformation of manufacturing has gradually spread from the eastern coastal areas to the central and western inland areas. (2) In horizontal comparison, among the five types of "ABCDE" digital modules constructed, artificial intelligence develops the fastest, cloud computing index value is second, and block chain value is the smallest. In vertical comparison, the leading provinces such as Beijing, Guangdong, and Shanghai have certain stability and a solid leading position, and there are occasional highlights in the central and western provinces. (3) In terms of polarization distribution, the digitalization of the manufacturing industry has obvious multi-peak patterns, showing the phenomenon of multi-polarization of digital services. The eastern region has both aggregate advantages and equilibrium disadvantages. (4) In terms of industry differences, the level of digital transformation in the high-end manufacturing industry is significantly higher than that in the mid-end and low-end industries. On the ownership attributes of enterprise digital transformation, private enterprises are the highest, followed by foreign-funded enterprises, and state-owned enterprises are the lowest. This research provides theoretical enlightenment and factual reference for manufacturing enterprises to carry out digital activities.