Visualizing Africa in Chinese Media: A Preliminary Computer-Assisted Study of 1950s-1980s Representation in Journal Illustrations and Book Covers

This poster presents the first phase of an ongoing research project examining the visual representation of Africa in Chinese journal illustrations and book covers from the 1950s to the 1980s. In this initial stage, we analyze visual materials, including 49 Chinese-language books on Africa as well as 23 issues of the official English-language journal of the Afro-Asian Writers’ Association, whose secretariat was based in Beijing. Using computer vision techniques, we identify key visual elements—such as people, symbols, and landscapes—featured in these publications. After detecting these elements, machine learning methods, along with manual verification, are applied to identify recurring patterns and themes, enabling cross-validation between human analysis and automated results. This combined approach strengthens the identification of visual motifs and links them to broader historical themes, such as ‘Third World’ cosmopolitanism, which transcended national boundaries and the Cold War binary.

Findings from this phase provide preliminary insights into how Africa was visually represented in Chinese media and demonstrate the feasibility of using computational tools and machine learning models to extract, categorize, and analyze visual elements in historical archival materials in Afro-Asian studies. Subsequent phases will continue employing computer-assisted techniques, with more advanced analyses such as stylistic analysis, temporal analysis to trace the evolution of representations over time, and cross-media comparisons to explore broader visual trends. In future stages, the project may expand to include a wider range of visual materials, such as magazines, posters, and films.

Poster presentation is co-authored with: Fudie Zhao – DPhil candidate in the Faculty of Asian and Middle Eastern Studies, University of Oxford