Evaluating Actors’ Behavior on Social Media

Abstract

While social networks have become primary promotion platforms for TV series, it is crucial to provide reliable measurements of promotion effectiveness for actors, which can guide them to select better promotion strategies when they post microblogs. In this paper, influence indexes are proposed to measure the influence of microblogs, and some measurements on actors’ microblogs also indicate and reveal some useful patterns in their promotion behaviors. Then, a propensity score-matching method is applied to these data to identify effective promotion strategies at two promotion periods. In experiments, the proposed model is shown to be significant by [Formula: see text]-test evaluation and the model is demonstrated to be adequately specified by balance diagnostics. We also do sensitivity analysis to analyze how hidden covariates impact conclusions. With this application of microblog data, the causal effects between promotion strategies and promotion results can be assessed. Finally, we treat not only baseline covariates, but also the promotion strategies as features and predict the influence of promotion microblogs. With these features, we can acquire a more accurate prediction effect.

The role and influence of perceived experts in an anti-vaccine misinformation community

“…We can then compare the average influence between perceived experts and perceived non-experts while adjusting for confounding from covariates that may predict being a perceived expert without being a direct consequence of perceived expertise. Propensity score matching has been applied previously to compare the effectiveness of different users in the context of online advertising [ Yan et al, 2018 , Kuang et al, 2018 ] and to understand whether following prominent public health institutions on Twitter impacts vaccine sentiment [ Rehman et al, 2016 ]. Propensity score matching was performed using the MatchIt package in R [ Ho et al, 2011 ].…”