Robustness Test of Explanatory Variables in Intelligent Development Evaluation

In order to conduct a robustness test, explanatory variables were substituted. Specifically, the entropy weight method was employed in the benchmark regression of the comprehensive evaluation system for intelligent development, and both vertical and horizontal methods were utilized. The resulting regression outcomes are presented in columns (1) and (2) of Table 4. Additionally, the development of enterprise-level intelligence was investigated through the 'Enterprise Query' credit information platform, using Python crawler technology to search for relevant keywords such as 'intelligence', 'cloud', 'data', 'Internet', and 'machine learning' in the 'business scope' and 'company name' columns. The number of intelligent enterprises in each city was then used to gauge the degree of intelligent development, with the corresponding results reported in columns (3) and (4).