Reproduce “Dataset of Propaganda Techniques of the State-Sponsored Information Operation of the People's Republic of China” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Rong-Ching Chang, Chun-Ming Lai, Kai-Lai Chang, Chu-Hsing Lin
The authors introduce a dataset labeling rhetorical techniques in Mandarin tweets sampled from a Twitter-disclosed PRC state-backed information-operations batch. Unlike a yes-or-no propaganda judgment, its multi-label annotations identify devices such as Whataboutism and Loaded Language. They fine-tune Chinese BERT to classify these labels, reporting 0.85431 micro-F1 and 0.20803 macro-F1 on an 80/20 split, with no comparison or test beyond this dataset.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.