摘要
The innovative, uncertain and external nature of the fishing industry dictates that industrial policy plays an important role in its development. Rapid development through policy support has led to excellent results and leadership, while contributing to China's economic growth. Using text mining techniques, this paper conducts a hierarchical association mining of China's fishery policies based on three dimensions of policy keyword co-occurrence, policy themes and policy tools, and employs visualization methods to better reveal the dynamic evolutionary patterns of policies. Based on 5311 fishery policy texts at the national, provincial and municipal levels from 2003 to 2022 in China, three techniques of co-occurrence analysis, LDA topic modelling and similarity calculation were used to mine the texts respectively, and CiteSpace, LDAvis and ThemeRiver methods were used to visualize and analyze them. The following conclusions were obtained: (1) In terms of co-occurrence of policy keywords, there are patterns of "progressive evolution", "hierarchical lag" and "hierarchical lagged evolution". (2) In terms of policy themes, there is an evolutionary development of industries, enterprises, technologies and products from the macro to the micro level, and the "hierarchical stage correlation" feature is more obvious. (3) In terms of policy instruments, there are obvious "environmental policies are the most numerous, demand-based policies are the second most numerous, and supply-based policies are the least numerous", "the hierarchy lags behind development", "the role of central policies is strong, but the role of local policies is weak ", and "emphasis on project construction, but not on public services". Further, this paper puts forward policy recommendations in terms of shortening the time lag of industrial policies, strengthening policy coordination and improving policy content, aiming to improve the market-oriented mechanism of fishery policies, optimise public services and promote development.