The mechanism of innovation-driven emerging technology generation based on big data fusion in the perspective of technological self-reliance and self-improvement

Abstract

Emerging technologies create opportunities for later adopters to achieve technological and economic leapfrogging and are highly valued by governments and enterprises. In this paper, we first screen the key factors influencing innovation drive through data region feature extraction to achieve a synergistic innovation effect between collaborative innovation and R&D subjects. Secondly, the multidimensional data are fused, and the feature extraction is performed using Transformer’s encoder (Encoder) structure, and the bidirectional coding of the input sequence text is realized using the supporting MLM training method. Finally, by comparing the analysis with other multimodal fusion methods in the constructed real dataset, the high performance of this method on emerging technology innovation-driven problems is demonstrated. The experimental results show that the absolute path coefficient of the innovation environment on collaborative innovation capability is 0.728 and the standardized coefficient is 0.835, which indicates that the innovation environment has a significant positive correlation with the innovation capability of emerging technology R&D subjects. The innovation-driven performance of science and technology emerging technology generation mechanism based on big data fusion technology is improved by 34.2%. The innovation-driving model based on big data fusion technology proposed in this paper plays a positive role in promoting the agglomeration of emerging industries and effectively improves the innovation ability and the conversion rate of R&D results of emerging enterprises, which is of great strategic significance for future economic development.

Keywords

• emerging technologies
• innovation-driven
• big data fusion technologies
• feature extraction
• multimodality
• 62-07