Abstract:

The paper proposes a wordVAE digital humanities knowledge production feature

mining method based on the combination of word2vec and Variational Auto-Encoder (VAE).

The study conducts an empirical research on the characteristics of knowledge production taking

digital humanities research results (CSSCI Journal Papers) and related books as research

objects. The research shows that ：in terms of the research results of Journal Papers, the main

characteristics of digital humanities knowledge production are the coexistence of knowledge

production structure and boundary blurring, and interdisciplinary collaboration. In terms of

the research results of books, the main characteristics are thematic cohesiveness and broad

horizons.

Key words:

Digital humanities Knowledge production characteristics WordVAE

Paradigm shift

CLC Number: 

• G237