Perceived or Not Perceived: Film Character Models for Expressive NLG

Abstract

Abstract. This paper presents a method for learning models of character linguistic style from a corpus of film dialogues and tests the method in a perceptual experiment. We apply our method in the context of SpyFeet, a prototype role playing game. In previous work, we used the PERSONAGE engine to produce restaurant recommendations that varied according to the speaker's personality [14,12]. Here we show for the first time that: (1) our expressive generation engine can operate on content from the story structures of an RPG; (2) PERSONAGE parameter models can be learned from film dialogue; (3) PERSONAGE rule-based models for extraversion and neuroticism are be perceived as intended in a new domain (SpyFeet character utterances); and (4) that the parameter models learned from film dialogue are generally perceived as being similar to the character that the model is based on. This is the first step of our long term goal to create off-theshelf tools to support authors in the creation of interesting dramatic characters and dialogue partners, for a broad range of types of interactive stories and role playing games.

A Robot’s Expressive Language Affects Human Strategy and Perceptions in a Competitive Game

“…where the probabilities on the right-hand side are calculated according to (1), V gives the AFINN affective valence of a word (or 0 if not in the dictionary), A indicates whether the affect is encouraging (+1) or discouraging (-1), and the z i values are weights. We train our model on transcripts of popular films from the IMSDb archive [34], [35]. The code to generate the model from any corpora and make predictions based on arbitrary sentence stems can be found on Github.…”

Proceedings of the Second Workshop on Storytelling

“…For example, Storybook (Callaway and Lester, 2002), the Narrator (Theune et al, 2007) and Curveship (Montfort, 2009) all generate story text from an underlying fabula. Other research focused on aspects of the telling of the story, for example stylistic variation (Montfort, 2007), affective language (Strong et al, 2007) and personality (Lukin et al, 2014;Walker et al, 2011).…”

Strategic Talk in Film

“…Methods developed and evaluated on non-fictional texts were also applied to the analysis of fictional dialogues (Kundu, Das, and Bandyopadhyay 2013; Gorinski and Lapata 2015; Shen 2011; Karsdorp et al 2015). Although limited to linguistic style analysis, Walker et al (2011) showed that the detection of some of the communicative patterns and linguistic features of a character can be achieved (Walker et al 2011). Specifically, the published approach relied on a set of linguistic features, aggregated and extracted jointly from all dialogues a given character conducted, without differentiating conversational partner, point in the dramatic arc, or narrative relevance.…”