TY - JOUR
T1 - Emotion-based cross-variety irony detection
AU - Calvo, Hiram
AU - Gambino, Omar Juárez
N1 - Publisher Copyright:
© 2019 CEUR-WS. All rights reserved.
PY - 2019
Y1 - 2019
N2 - This work is centered on the data made available for the IroSvA challenge, consisting of three variants of Spanish language from three different countries. We propose a simple model for identifying irony, based on tweet embeddings, refraining from using of additional NLP techniques. We aim to find cues that are able to generalize the knowledge obtained from a language variant, and evaluate the ability to detect irony in different combinations of variants, from different countries and topics. For this purpose, we propose using six features based on the degree of emotion present in each tweet. These automatically tagged features include 5 levels of strength, ranging from none to very high, of six emotions: love, joy, surprise, sadness, anger, and fear. Experiments were carried out with different combinations of language variants. Obtained results show that exclusively using the information of the emotion levels (discarding the embeddings) could improve the irony detection in a language variant different from that used for training.
AB - This work is centered on the data made available for the IroSvA challenge, consisting of three variants of Spanish language from three different countries. We propose a simple model for identifying irony, based on tweet embeddings, refraining from using of additional NLP techniques. We aim to find cues that are able to generalize the knowledge obtained from a language variant, and evaluate the ability to detect irony in different combinations of variants, from different countries and topics. For this purpose, we propose using six features based on the degree of emotion present in each tweet. These automatically tagged features include 5 levels of strength, ranging from none to very high, of six emotions: love, joy, surprise, sadness, anger, and fear. Experiments were carried out with different combinations of language variants. Obtained results show that exclusively using the information of the emotion levels (discarding the embeddings) could improve the irony detection in a language variant different from that used for training.
UR - http://www.scopus.com/inward/record.url?scp=85071183145&partnerID=8YFLogxK
M3 - Artículo de la conferencia
AN - SCOPUS:85071183145
SN - 1613-0073
VL - 2421
SP - 264
EP - 271
JO - CEUR Workshop Proceedings
JF - CEUR Workshop Proceedings
T2 - 2019 Iberian Languages Evaluation Forum, IberLEF 2019
Y2 - 24 September 2019 through 24 September 2019
ER -