Loving and hating the movies in English, German and Spanish

Abstract

We present a quantitative analysis of evaluative language in a genre in which it is particularly prominent, that of movie reviews. The data chosen are non-professional consumer-generated reviews written in English, German and Spanish. The reviews are analysed in terms of the categories of Attitude and Graduation within the Appraisal framework (Martin and White, 2005). A number of similarities in the distribution of the Appraisal subcategories were found across the three languages, such as the high frequency of Appreciation and the narrow relationship between the global polarity of the reviews and the individual polarity of the spans. More importantly, the analysis uncovers a number of cross-linguistic distributional differences, which may be explained in terms of a wide array of factors, such as lexicogrammar, word order, argumentative style or sociocultural reasons.

The SFU Opinion and Comments Corpus: A Corpus for the Analysis of Online News Comments

“…As in the negation annotations, the comments were annotated by two individuals (one an author of this paper; the other a research assistant), evaluated for agreement, and curated to increase the accuracy of annotation. We began with guidelines previously used on a corpus of film reviews (Taboada et al 2014) and the step-wise process in Fuoli (2018) and further developed them by iteratively testing and discussing those guidelines on our corpus. Once guidelines had been established, a research assistant annotated the original 1121 comments under the supervision of one of the researchers, who was responsible for curating the annotations.…”

“…For our annotations, we found inspiration in existing work on manual annotation for Appraisal (Whitelaw et al 2005;Read and Carroll 2012;Taboada et al 2014), including the set of steps in Fuoli (2018) and Fuoli and Hommerberg (2015). The annotations consist of two layers: Attitude and Graduation.…”

Using Rhetorical Structure Theory for Detection of Fake Online Reviews

“…This strategy of expressing 'caveats' has been noted as a feature of negative English language movie reviews (Taboada et al, 2014). Hedged positive evaluation has also been found to feature in Japanese academic book reviews (Itakura, 2013).…”

Building an annotated dataset of app store reviews with Appraisal features in English and Spanish

“…Attitude is concerned mainly with feelings, such as emotions, judgements and evaluations; it can be further subdivided into Affect, Judgement and Appreciation, each of which is subdivided into more delicate categories, as shown in Figure 1. (Mora 2011, Carretero and Taboada 2009, 2010a, 2010b, 2014 and other text types (Taboada, Carretero and Hinnel, 2014;Lavid et al 2014;Lavid, Carretero and Zamorano 2016).…”