Digital humanities and archival studies: a cross-linguistic analysis

Abstract

This study aims to examine how digital humanities methodologies, particularly natural language processing and network analysis, have influenced archival scholarship across Chinese- and English-language contexts from 2004 to 2023. The study uses latent Dirichlet allocation for topic modeling, term frequency-inverse document frequency for keyword extraction and social network analysis to examine thematic patterns in archival journal articles. Chinese knowledge information processing tagger is used for Chinese-language data, while Gensim with Chinese and Gensim natural language toolkit is applied to English-language data, enabling cross-linguistic comparison of thematic patterns. The results reveal distinct thematic divergence. Chinese-language journals emphasize state-led initiatives, such as national identity construction and digital infrastructure, whereas English-language journals focus on community archives, Indigenous rights and archival justice. Despite these differences, both corpora converge on themes such as digital archive management, policy dissemination and archival education, reflecting a shared emphasis on archival infrastructure and governance. Topic modeling identifies six coherent themes in Chinese articles and seven in English. Keyword analysis shows that Chinese literature prioritizes institutional roles, whereas English texts emphasize social justice and community engagement. The study’s primary contribution lies in its cross-linguistic comparative design, which integrates topic modeling, keyword clustering and social network analysis across Chinese- and English-language scholarly corpora. This approach highlights divergent orientations in archival knowledge production across the two communities, a dimension that has received limited systematic attention in prior single-language or single-method studies.