Investigation of Goodreads’ reviews: Kakutanied, deceived or simply honest?

Abstract

Purpose The purpose of this paper is to present an analysis of Goodreads’ user-generated book reviews from a linguistic perspective for insights into the psychological aspects of reviewers’ perceptions and behaviors. This examination of users’ language and perspectives may shed light on the role and value of user-generated reviews in complementing the traditional representation of resources and facilitating the discoverability of cultural objects. Design/methodology/approach This study involved a textual analysis of 474,803 unique reviews of Goodreads’ 2015 top-rated books generated by 9,335 Goodreads’ reviewers. In order to better understand the nuances of user-generated reviews, a content analysis was applied to 2,500 reviews of each of the five top-ranked titles in Goodreads’ Fiction Literature genre category. Findings The analysis of user-generated reviews demonstrates that language is a quite stable and reliable dimension across Goodreads’ users. The high rate of function words utilized, in particular I-words, coupled with positive emotion words, suggests that reviewers tended to convey their opinions in order to influence other individuals’ reading choices, or in Bourdieu’s (1985) terms, influence cultural production. In line with previous studies of user-generated reviews, the prevalence of positive reviews may also imply their unreliable nature. This study supports the importance of transparency regarding inclusion of user-generated reviews in traditional systems of knowledge representation, organization and discovery, such as WorldCat. Originality/value This study contributes to better understanding of linguistic characteristics of Goodreads’ reviews, including the role and value of user-generated reviews in complementing traditional representation of resources and facilitating discoverability of cultural objects.

An investigation of cultural objects in conflict zones through the lens of TripAdvisor reviews: A case of South Caucasus

“…In line with previous studies of user-generated reviews [49], the reviews analysed in this study were predominantly positive: 1002 (88%) of Abkhazian and 161 (83.4%) of Nagorny Karabakh–related reviews, followed by neutral 110 (9.7%) and 25 (13%), and negative 26 (2.3%) and 8 (4.1%), respectively. This could be explained by the overall positive nature of user-generated reviews [50] as well as visitors’ conscious choices and preferences in exploring particular cultural sites.…”

Modelling social readers: novel tools for addressing reception from online book reviews

“…Social reading platforms offer an opportunity to explore these aspects of reader response and narrative structure outside of the laboratory or classroom setting [ 1 , 2 , 4 , 32 – 34 ]. Work on these platforms has focused on an ethnography of online reading [ 35 ], considerations of reader review sentiment [ 36 ], how gender and intimacy are intertwined in reviews [ 37 ] and broader linguistic characteristics of the reviews [ 38 ]. Other studies have focused on Goodreads in the context of socially networked reading [ 39 ], the moderation of reader responses [ 40 ], recommender systems and participants’ motivations for reviewing [ 41 ], grassroots genre classification [ 3 ] and prediction of the success of a book given certain features in the reviews [ 42 ].…”

Investigating the effects of Goodreads challenges on individuals reading habits

“…The content of reviews are also extensively studied [10,30,31]. [15] investigates the lingual features of reviews, demonstrating that user-generated reviews are unreliable as they are mostly positive and attempt to persuade others to read the book as well. [33] use aggregate reviews as a means for story-graph creation and find their method to be quite accurate compared to the book's true network.…”