Dear Mergen,

Very interesting article on the meta-analysis of the model, it is relevant for the historical context, The model needs to be updated to include new confounding factors that may exert significant influences on the length and features of each phase and newly discovered phases. Search engines, Facebook, Twitter, Wiki-Leaks, other newly formed professional social media like this, including AI social bots are disrupting the knowledge transfer and accessibility of information and misinformation. We may need deep learning or machine learning models to understand and predict the ramifications of new public policies. 

Yours sincerely,

Keith

Keith,

Thanks for your reply. Indeed, the meta-analysis refers to the importance of internet tools, e.g. search engines and social media, in modern agenda-setting. I still generally refer to the role of internet rather than to social media in particular, since the latter cannot always precisely trace public (netizen) attention due to constant migration factors. For instance, relying on Twitter these days might be especially risky (who actually uses Twitter these days, and is it really popular and representative of wider populace?)

I agree on the general importance of knowing the nuances of machine learning. However, it is the human that should stay in control, i.e. interpreting and making predictions on new policy avenues.

Mergen