Scholarlink

This study explores the impact of Artificial Intelligence Technology (AIT) and Big Data Technology (BDT) on firms’ Digital Innovation Capability (DIC). Using the data from publicly listed companies, the multiple regression analysis and machine learning methods (Decision Tree, Random Forest, and Gradient Boosting) are employed to construct regression models, comparing predictive performance across algorithms and analyzing feature importance and partial dependence plots. The results indicate that AIT has a significant positive effect on DIC, whereas BDT shows a more complex impact, exhibiting a negative effect, particularly when investment is excessive, which may inhibit innovation. Partial dependence plot analysis reveals that AIT has a substantial marginal contribution to DIC at initial investment stages, which gradually diminishes; conversely, BDT’s marginal contribution demonstrates an overall declining trend. This study provides empirical evidence for firms’ technology investment decisions in the digital innovation domain and offers recommendations for optimizing AIT and BDT applications to enhance digital innovation capabilities.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The Interaction Effect of AI and Big Data on Corporate Digital Innovation: An Empirical and Machine Learning Approach

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Yang Lu

摘要

This study explores the impact of Artificial Intelligence Technology (AIT) and Big Data Technology (BDT) on firms’ Digital Innovation Capability (DIC). Using the data from publicly listed companies, the multiple regression analysis and machine learning methods (Decision Tree, Random Forest, and Gradient Boosting) are employed to construct regression models, comparing predictive performance across algorithms and analyzing feature importance and partial dependence plots. The results indicate that AIT has a significant positive effect on DIC, whereas BDT shows a more complex impact, exhibiting a negative effect, particularly when investment is excessive, which may inhibit innovation. Partial dependence plot analysis reveals that AIT has a substantial marginal contribution to DIC at initial investment stages, which gradually diminishes; conversely, BDT’s marginal contribution demonstrates an overall declining trend. This study provides empirical evidence for firms’ technology investment decisions in the digital innovation domain and offers recommendations for optimizing AIT and BDT applications to enhance digital innovation capabilities.