Visualization of the Semantic Knowledge Landscape of Editing and Publishing Domain in China

The rapid advancement of information technology is reforming the ecosystem of the editing and publishing industry. Under the demand for building novel publishing patterns, it is critical to map out the research trends and evolutionary trajectories of the field of editing and publishing comprehensively. In this study, drawing on large-scale journal papers, we construct the knowledge landscape of editing and publishing research field in China, by manifold learning algorithm UMAP based on the deep semantic associations between papers learned by Doc2vec, to visualize the static and diachronic structure of this field. Firstly, with the Gaussian kernel density function to characterize the heterogeneity of spatial distribution of papers, we identify core research topics in the editing and publishing domain. Then, by respectively constructing cumulative and sliced dynamic knowledge maps, we find that over the past 40 years, the research scope has continued to expand, evidenced by the emergence of new research topics along the edges of the map, and meanwhile, topics within the field keep merging and fusing. Furthermore, according to the pattern of research hotspot transitions, the development is roughly divided into four stages. These findings offer valuable insights for researchers in the publishing field and scientific policymakers.