Grounding Characters and Places in Narrative Text

37979

Authors

nan

Abstract

AbstractTracking characters and locations throughout a story can help improve the understanding of its plot structure. Prior research has analyzed characters and locations from text independently without grounding characters to their locations in narrative time. Here, we address this gap by proposing a new spatial relationship categorization task. The objective of the task is to assign a spatial relationship category for every character and location co-mention within a window of text, taking into consideration linguistic context, narrative tense, and temporal scope. To this end, we annotate spatial relationships in approximately 2500 book excerpts and train a model using contextual embeddings as features to predict these relationships. When applied to a set of books, this model allows us to test several hypotheses on mobility and domestic space, revealing that protagonists are more mobile than non-central characters and that women as characters tend to occupy more interior space than men. Overall, our work is the first step towards joint modeling and analysis of characters and places in narrative text.

Keywords

nan

Full Content

Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics Volume 1: Long Papers, pages 11723–11736 July 9-14, 2023 ©2023 Association for Computational Linguistics Grounding Characters and Places in Narrative Texts Sandeep Soni and Amanpreet Sihra University of California, Berkeley sandeepsoni,amisihra@berkeley.edu Elizabeth F. Evans Wayne State University e.f.evans@wayne.edu Matthew Wilkens Cornell University wilkens@cornell.edu David Bamman University of California, Berkeley dbamman@berkeley.edu Abstract Tracking characters and locations throughout a story can help improve the understanding of its plot structure. Prior research has analyzed char- acters and locations from text independently without grounding characters to their locations in narrative time. Here, we address this gap by proposing a new spatial relationship cate- gorization task. The objective of the task is to assign a spatial relationship category for ev- ery character and location co-mention within a window of text, taking into consideration lin- guistic context, narrative tense, and temporal scope. To this end, we annotate spatial relation- ships in approximately 2500 book excerpts and train a model using contextual embeddings as features to predict these relationships. When applied to a set of books, this model allows us to test several hypotheses on mobility and domestic space, revealing that protagonists are more mobile than non-central characters and that women as characters tend to occupy more interior space than men. Overall, our work is the first step towards joint modeling and analy- sis of characters and places in narrative text. 1 Introduction The association between characters and the places they navigate is central to a wide range of literary phenomena: Bildungsromane depict a character’s journey across geographic space as a component of their psychological coming of age (Bakhtin, 1987; Jeffers, 2016); the flâneur, who walks and observes throughout a city, epitomizes the power that fol- lows from peripatetic access to public spaces (Ben- jamin, 2002; Wolff, 1985; Wilson, 1992); class, gender, and racial associations can render surpris- ing or scandalous a character’s mere presence in an otherwise innocuous location. While much work in literary history and theory has explored this interaction, it has remained out of reach for empirical observation and large-scale comparisons with social constructs such as gender, Mr. Fogg repaired to [his bedroom]TO, and called out, “Passepartout!” As for Mr. Fogg, he slumbered as peacefully as if he had been in [his serene mansion in [Saville Row]NO REL ]NO REL. On the ninth day after leaving [Yokohama]FROM, Phileas Fogg had traversed exactly one half of the terrestrial globe . Phileas Fogg wished to go to [Liverpool]NO REL , but the captain would not carry him [there]NO REL . Last Friday at eight o'clock in the evening, Monsieur Fogg came [home]TO from [his club]FROM Figure 1: In Around the World in 80 Days, Phileas Fogg is mentioned in conjunction with a wide range of places, but is only physically grounded in some. Disentangling the places he is in from the places he is not is crucial for tracking his movement throughout this work. . social status (Cresswell, 2012) and agency (Sen, 1993). In …

Content truncated to 500 words

Labeled Text

AbstractTracking characters and locations throughout a story can help improve the understanding of its plot structure. Prior research has analyzed characters and locations from text independently without grounding characters to their locations in narrative time. Here, we address this gap by proposing a new spatial relationship categorization task. The objective of the task is to assign a spatial relationship category for every character and location co-mention within a window of text, taking into consideration linguistic context, narrative tense, and temporal scope. To this end, we annotate spatial relationships in approximately 2500 book excerpts and train a model using contextual embeddings as features to predict these relationships. When applied to a set of books, this model allows us to test several hypotheses on mobility and domestic space, revealing that protagonists are more mobile than non-central characters and that women as characters tend to occupy more interior space than men. Overall, our work is the first step towards joint modeling and analysis of characters and places in narrative text.

Publication Details

Paper ID:
37979
Year:
2023
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan