Home // HUSO 2016, The Second International Conference on Human and Social Analytics // View article

Producing Affective Language. Content Selection, Message Formulation, and Computational Modelling

Authors:
Martijn Goudbeek
Nadine Braun
Charlotte Out
Emiel Krahmer

Keywords: Emotion and Cognition; Language production; Referring expressions; Natural Language Generation.

Abstract:
We introduce a project investigating how emotional states influence language production using both experimental and corpus based approaches. Here, we illustrate our project by asking whether content selection (“deciding what to say”) and linguistic realization (“deciding how to say it”) are affected by the emotional state of a speaker. We do this first by assessing whether disgusted speakers are more or less prone to align with their dialogue partners than amused speakers. Second, we develop a corpus of emotionally laden soccer reports that, even though they refer to the same event, will differ depending on whether the report comes from the winning or losing team. In both cases, we focus on the production and analysis of referring expressions. Our findings will be used to build an affective natural language generation system.

Pages: 43 to 47

Copyright: Copyright (c) IARIA, 2016

Publication date: November 13, 2016

Published in: conference

ISSN: 2519-8351

ISBN: 978-1-61208-519-7

Location: Barcelona, Spain

Dates: from November 13, 2016 to November 17, 2016