Abstract: Based on the data of key publication project topics for theme publishing from 2012 to 2023, using Python software and Jieba text mining technology to perform frequency statistics on the keywords of long-term theme publishing project topics, this paper analyzes the development trend and era characteristics of theme publishing project topics. It has found that topic selection for theme publishing projects has shown three characteristics in recent years: firstly, multidimensional trend of the theme, which emphasizes not only the node dimension but also the macro and eternal dimensions; secondly, while focusing on the political nature of theme publishing, the perspective expands to all aspects of social life surrounding the central work; thirdly, while focusing on the theoretical aspects of theme publishing, various textual forms should be utilized to emphasize the acceptability of theme publishing. The conclusion provides certain reference value for the top-level design of theme publishing and the application of theme publishing projects by publishing houses.

Key words: Theme publishing　Project topic selection　Word frequency　Topic selection trend