Research Question: This study aims at uncovering China’s national propaganda strategies by analyzing texts from two sets of state media: Xinwen Lianbo and People’s Daily’s official WeChat public account.

Data: I used web-scraping to collect transcription of Xinwen Lianbo from the XWLB official digital record website and People’s Daily’s official WeChat public account blog.

Methods: I mainly used text analysis to classify and extract meaningful information from China’s propaganda platforms. The specific techniques I applied include keyword extraction, sentiment analysis, dictionary-based content analysis, supervised/unsupervised text classification, etc.​​​​​​​

Challenges: There are two challenges in processing propaganda texts, which are common in Chinese Natural Language Processing: 1. Word segmentation. Unlike Latin-derived languages, there is no space between the characters in Chinese, causing ambiguities in deciphering its meaning. 2. Derived from the first challenges, effective text analysis and machine learning algorithms applied to Chinese are also limited

Findings: The agendas of different state media platforms diverge significantly, with Xinwen Lianbo focusing more on politics and People’s Daily’s official WeChat public account putting more emphasis on social life. For future analysis, I am investigating more structured differences between and features of these two sets of state media.