Abstract
Does the People's Republic of China (PRC) interfere with European electionsthrough ethnic Chinese diaspora media? This question forms the basis of anongoing research project exploring how PRC narratives about European electionsare represented in Chinese diaspora media, and thus the objectives of PRC newsmedia manipulation. In order to study diaspora media efficiently and at scale,it is necessary to use techniques derived from quantitative text analysis, suchas topic modelling. In this paper, we present a pipeline for studyinginformation dynamics in Chinese media. Firstly, we present KeyNMF, a newapproach to static and dynamic topic modelling using transformer-basedcontextual embedding models. We provide benchmark evaluations to demonstratethat our approach is competitive on a number of Chinese datasets and metrics.Secondly, we integrate KeyNMF with existing methods for describing informationdynamics in complex systems. We apply this pipeline to data from five newssites, focusing on the period of time leading up to the 2024 Europeanparliamentary elections. Our methods and results demonstrate the effectivenessof KeyNMF for studying information dynamics in Chinese media and lay groundworkfor further work addressing the broader research questions.