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

Online reviews provide users with the opportunity to rate various types of items such as movies, music, and video games using a combination of numeric scores and textual comments. The study proposes a novel method that applies statistical matching on network-based covariates, with the aim to improve the estimation of the association between words and highly controversial items in online reviews. The application of this method on a sample of 40,665 items from the website Metacritic detects 218 highly controversial items. The application supports the theory that controversies on Metacritic are driven with a sense of self-awareness of participating of an online controversy (‘review bombing’). Typical controversial topics (sexual identities, religious morality, politics) are associated with controversial reviews, too.