Fig. 1

Our sketch of the (adaptive) community-response approach. Note that \(t\) and \(r\) stand for (story-supporting) tweet and reply, respectively. Here, tweets \(t_1, t_6\) and \(t_7\) can be identified via fact-checking (’seen’ tweets). However, the community-response approach fails for \(t_2, t_3, t_4, t_5, t_8\) and \(t_9\) (’unseen’ tweets), because there were no fact-checking responses