基于神经网络词转向量的新质生产力关注热点比较研究

New qualitative productivity is a contemporary interpretation of Marxist theory on productivity, representing advanced productivity driven by technological innovation and the multiplier effect of data elements. It has attracted widespread attention in both academic and practical circles. This paper compares and analyzes the focus on new qualitative productivity between China\'s academic platform (CNKI) and social media platforms such as Weibo and Zhihu, aiming to promote further achievements in research on new qualitative productivity in China. Firstly, abstracts related to new qualitative productivity were collected from CNKI, while responses and blog posts related to it were gathered from Weibo and Zhihu, forming corpora for the academic and social media platforms respectively. Then, the emerging neural network-based Word2vec algorithm was applied to these two corpora for modeling, enabling effective measurement of word similarity at both syntactic and semantic levels. On this basis, t-SNE dimensionality reduction technique was used to reduce these high-dimensional word vectors to two dimensions for visualization, comparing and analyzing the positional relationships of important words like "new qualitative productivity" in the semantic space. The research results indicate differences in the focus on new qualitative productivity between China\'s academic and social media platforms. This paper innovatively employs word vectors to compare and analyze the focus on new qualitative productivity between China\'s academic and social media platforms, providing a new perspective for research on new qualitative productivity.