A network-based matching design for text mining of hyper-polarised online reviews

Online reviews provide users with the opportunity to rate various types of items such as 3 movies, music, and video games using a combination of numeric scores and textual com4 ments. The study proposes a novelmethod that applies statisticalmatching on network-based 5 covariates,with the aimto improve the estimation of the association betweenwords and highly 6 controversial items in online reviews. The application of this method on a sample of 40,665 7 items from the website Metacritic detects 218 highly controversial items. The application 8 supports the theory that controversies onMetacritic are driven with a sense of self-awareness 9 of participating of an online controversy (‘review bombing’). Typical controversial topics 10 (sexual identities, religious morality, politics) are associated with controversial reviews, too.