“…Furthermore, our results might be skewed with respect to the timeframe of the study, which was just after the measles outbreak in California Disneyland and it would therefore be interesting to carry out the analysis over a more longitudinal timeframe and also without assuming each user (including each bot user) to have same opinion. However, work by Vraga et al (2017) on the discussion of cancer in social media showed peaks of traffic during awareness months and little in other time frames, which could be broadly considered similar to the disease outbreak studied here. In addition, as Bello-Orgaz et al (2017) found that influential users (i.e., those with high degree centrality) are often in pro-vaccine communities and Love, Himelboim, Holton, & Stewart (2013) discovered that no particular subject is dominating the vaccine conversation in Twitter, it would, therefore, be interesting for future research to explore whether "influential" users are more of those with high degree centrality or those who they agree, and thus understand what defines a user to be influential in the context of anti-vaccination or other highly divided online discussions (Cha, Haddadi, Benevenuto, & Gummadi, 2010).…”