Exploring emotions in online movie reviews for online browsing

A restaurant review is a reflection of the reviewer's experience and attitude towards the restaurant. The same applies to a review of a new phone or a review on any other online merchandize. Films, however, are created with the intended purpose to evoke an emotional response in the viewer. This emotional response does not necessarily correspond with the viewer's attitude towards the film. Thus, the question we try to address is, would the emotions expressed in a film's online reviews also reflect the emotions elicited during the film? In this work, we take a first step in the investigation of this question, by studying the role of emotions in movie reviews as expressed in a large dataset of millions of online reviews for over 9000 movies, that appeared in IMDb from 1972 to 2015. Our results show that we can extract emotions elicited by the film from its reviews, and create an emotional signature of a film, and of a genre. This is a first step towards an Emotion-based Film Browser UI system that will enable users to browse films according to the emotions they evoke. Copyright is held by the owner/author(s).