Towards Interactive Agents that Infer Emotions from Voice and Context Information

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

Conversational agents are increasingly being used for training of social skills. One of their most important benefits is their ability to understand the user`s emotions, to be able to provide natural interaction with humans. However, to infer a conversation partner’s emotional state, humans typically make use of contextual information as well. This work proposes an architecture to extract emotions from human voice in combination with the context imprint of a particular situation. With that information, a computer system can achieve a more human-like type of interaction. The architecture presents satisfactory results. The strategy of combining 2 algorithms, one to cover ‘common cases’ and another to cover ‘borderline cases’ significantly reduces the percentage of mistakes in classification. The addition of context information also increases the accuracy in emotion inferences.
LanguageEnglish
Article numbere2
Number of pages15
JournalEAI Endorsed Transactions on Creative Technologies
Volume17
Issue number10
DOIs
Publication statusPublished - 2017

Fingerprint

Computer systems

Cite this

@article{8faf3aebfac146bfae1a923d7e8e701f,
title = "Towards Interactive Agents that Infer Emotions from Voice and Context Information",
abstract = "Conversational agents are increasingly being used for training of social skills. One of their most important benefits is their ability to understand the user`s emotions, to be able to provide natural interaction with humans. However, to infer a conversation partner’s emotional state, humans typically make use of contextual information as well. This work proposes an architecture to extract emotions from human voice in combination with the context imprint of a particular situation. With that information, a computer system can achieve a more human-like type of interaction. The architecture presents satisfactory results. The strategy of combining 2 algorithms, one to cover ‘common cases’ and another to cover ‘borderline cases’ significantly reduces the percentage of mistakes in classification. The addition of context information also increases the accuracy in emotion inferences.",
author = "D. Formolo and Tibor Bosse",
year = "2017",
doi = "10.4108/eai.4-9-2017.153054",
language = "English",
volume = "17",
journal = "EAI Endorsed Transactions on Creative Technologies",
issn = "2409-9708",
number = "10",

}

Towards Interactive Agents that Infer Emotions from Voice and Context Information. / Formolo, D.; Bosse, Tibor.

In: EAI Endorsed Transactions on Creative Technologies, Vol. 17, No. 10, e2, 2017.

Research output: Contribution to JournalArticleAcademicpeer-review

TY - JOUR

T1 - Towards Interactive Agents that Infer Emotions from Voice and Context Information

AU - Formolo, D.

AU - Bosse, Tibor

PY - 2017

Y1 - 2017

N2 - Conversational agents are increasingly being used for training of social skills. One of their most important benefits is their ability to understand the user`s emotions, to be able to provide natural interaction with humans. However, to infer a conversation partner’s emotional state, humans typically make use of contextual information as well. This work proposes an architecture to extract emotions from human voice in combination with the context imprint of a particular situation. With that information, a computer system can achieve a more human-like type of interaction. The architecture presents satisfactory results. The strategy of combining 2 algorithms, one to cover ‘common cases’ and another to cover ‘borderline cases’ significantly reduces the percentage of mistakes in classification. The addition of context information also increases the accuracy in emotion inferences.

AB - Conversational agents are increasingly being used for training of social skills. One of their most important benefits is their ability to understand the user`s emotions, to be able to provide natural interaction with humans. However, to infer a conversation partner’s emotional state, humans typically make use of contextual information as well. This work proposes an architecture to extract emotions from human voice in combination with the context imprint of a particular situation. With that information, a computer system can achieve a more human-like type of interaction. The architecture presents satisfactory results. The strategy of combining 2 algorithms, one to cover ‘common cases’ and another to cover ‘borderline cases’ significantly reduces the percentage of mistakes in classification. The addition of context information also increases the accuracy in emotion inferences.

U2 - 10.4108/eai.4-9-2017.153054

DO - 10.4108/eai.4-9-2017.153054

M3 - Article

VL - 17

JO - EAI Endorsed Transactions on Creative Technologies

T2 - EAI Endorsed Transactions on Creative Technologies

JF - EAI Endorsed Transactions on Creative Technologies

SN - 2409-9708

IS - 10

M1 - e2

ER -