Innovation and Practice of the Ethical Supervision Mechanism for Computer-Assisted Publishing under the Integrated Publishing Model

Abstract

With the development of artificial intelligence, the ethical regulation of computer-aided publishing is strongly challenged. This paper establishes the process of ethical regulation of computer-aided publishing, calculates the text similarity, and uses the K-Means algorithm to cluster and analyze the subject keywords. The LDA model is established on the basis of Dirichlet distribution to mine the subject features of computer-aided publishing, and the classification key results are based on plain Bayes, so as to discuss the ethical misconduct of computer-aided publishing. The results show that 73.62% of the keywords in the field of computer-aided publishing, which accounted for the largest percentage, belong to verbs. The mean value of similarity of near-synonyms under publishing keywords reaches 0.864, which helps to sort out the association of different attribute word classes. Also the dendrogram of keywords obtained by clustering has the highest distance of 1.15. The computer-aided publishing regulatory mechanism established in this paper helps to adapt to the convergent publishing model.