Data elements have become the most characteristic production factors of the times, and their value release depends on whether they can be effectively circulated. This article divides the stock of data elements into data element generation volume and data element transaction volume, and incorporates them into a dynamic stochastic general equilibrium model that includes disruptive innovation. The N-gram model part of speech template, big language model semantic analysis, and word frequency inverse text frequency index/point mutual information method are used to extract technical terms. The technical features are verified through the Wikipedia entry directory, and the disruptive innovation of enterprises is calculated. Then, using the interleaved double difference model, we explored how the marketization of data elements can reconstruct the existing technological innovation paradigm. The results indicate that the marketization of data elements reduces data transaction friction, expands the proportion of enterprise data resources converted into effective data elements, and slows down the phenomenon of falling into data element traps due to the original accumulation of data elements. This process mainly works through two paths: data element quality and data element marginal productivity. To ensure the effective utilization of innovative effects in the marketization of data elements, the key lies in strengthening the supply of high-quality data and consolidating the construction of digital infrastructure. This article provides a new mechanism for incentivizing technological innovation paths and offers new insights for improving the data element market.