The next speaker at Web Science 2016 [4] is Onur Varol, who points out the wide variety of purposes for which people use social media, and notes that we change our online persona and usage styles according to different communicative contexts. Can we match language style and user intent, then?

The project's experiments found that messages written using logical arguments are perceived as more authoritative, while an absence of logical content makes the sender appear more likeable; the communication style also varies across different fields of interest (products, health, politics).

URL cascades tend to be more likely to involve relatively similar users. Sorry, I'm missing some detail here, because he's really not explaining his graphs sufficiently... Ultimately, simulations of heterogeneous-intent networks can create realistic cascades.