What are Chinese talking about in hot weibos?

SinaWeibo is a Twitter-like social network service emerging in China recently. We analyzed the hot weibos (tweets), which exceed threshold of being reposted for 1000 times, from a data set of 650 million weibos during August 2009 and January 2012. We classified the hot weibos into eight categories, namely Entertainment & Fashion, Hot Social Events, Leisure & Mood, Life & Health, Seeking for Help, Sales Promotion, Fengshui & Fortune and Deleted weibos. There are several findings. Firstly, Leisure & Mood and Hot Social Events account for almost 65% of all the hot weibos. This may indicate a potential dual-structure of the current society of China: On the one hand, economy of the country as a whole is gaining sustaining growth, which enables people to enjoy a better life and spare more time on leisure and mood topics. On the other hand, there still exist considerable amount of serious social problems, such as government corruption and environmental pollution, which draw people's concern and worries all the time. Secondly, users' posting and reposting behaviors are associated with user profiles, namely: (1) Gender. Male users generate two thirds of hot weibos. (2) Verification status. Verified users contribute 46.5% of hot weibos, who comprise only 0.1% in SinaWeibo user population. Interestingly, 39.2% of the verified-usergenerated weibos are written by SPA users (who generate weibos of a particular style, or in a consistent way, e.g. to say words of wisdom, "chicken-soup-soul" like sentences, and jokes etc.). This complements the previous finding of Yu et al. (2012), implying that SinaWeibo is in an "artificial inflation" not only on the reposting side but also on the posting side. Unfortunately, only 14.4% of the hot weibos are created by grassroots (not verified users). (3) Geographical location. Users from different areas of China show distinct posting and reposting behaviors, which partially reflect their indigenous cultures. Finally, homophily is also examined for people's reposting behaviors.