Abstract:In order to achieve a scientific evaluation of enterprise innovation capabilities and reveal the economic consequences of enterprise innovation, this paper selects sample data from 2005 to 2020 of listed companies in China's information industry, conducts text mining on patent text abstracts based on the TF-IDF algorithm weighted method, and constructs an evaluation system for enterprise innovation capabilities by combining detailed indicators under multiple dimensions. The relationship characteristics between enterprise innovation capabilities and enterprise performance and market value fluctuations are examined through empirical tests. The study found that: First, the enterprise innovation capability evaluation system constructed based on patent text information mining is scientific and reliable, and the innovation capability of the information industry presents a U-shaped feature, which is in line with the development trend of the information industry in recent years. Second, the enterprise innovation capability has a significant positive impact on enterprise performance. Third, based on the impact of innovation capability on enterprise performance, further research found that innovation capability has a significant positive impact on the improvement of enterprise market value. Fourth, through long-short portfolio testing and GRS testing, it was found that the addition of innovation capability factors can better explain stock returns. Therefore, the research not only enriches the theoretical research results of relevant enterprise innovation, but also provides theoretical guidance and experience basis for enterprises to achieve performance and market value growth by improving their own innovation ability.