CoAStaL at SemEval-2019 Task 3: Affect Classification in Dialogue using Attentive BiLSTMs
Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Standard
CoAStaL at SemEval-2019 Task 3 : Affect Classification in Dialogue using Attentive BiLSTMs. / González, Ana Valeria; Petrén Bach Hansen, Victor; Bingel, Joachim; Søgaard, Anders.
Proceedings of the 13th International Workshop on Semantic Evaluation. Association for Computational Linguistics, 2019. s. 169-174.Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - CoAStaL at SemEval-2019 Task 3
AU - González, Ana Valeria
AU - Petrén Bach Hansen, Victor
AU - Bingel, Joachim
AU - Søgaard, Anders
PY - 2019
Y1 - 2019
N2 - This work describes the system presented by the CoAStaL Natural Language Processing group at University of Copenhagen. The main system we present uses the same attention mechanism presented in (Yang et al., 2016). Our overall model architecture is also inspired by their hierarchical classification model and adapted to deal with classification in dialogue by encoding information at the turn level. We use different encodings for each turn to create a more expressive representation of dialogue context which is then fed into our classifier.We also define a custom preprocessing step in order to deal with language commonly used in interactions across many social media outlets. Our proposed system achieves a micro F1 score of 0.7340 on the test set and shows significant gains in performance compared to a system using dialogue level encoding.
AB - This work describes the system presented by the CoAStaL Natural Language Processing group at University of Copenhagen. The main system we present uses the same attention mechanism presented in (Yang et al., 2016). Our overall model architecture is also inspired by their hierarchical classification model and adapted to deal with classification in dialogue by encoding information at the turn level. We use different encodings for each turn to create a more expressive representation of dialogue context which is then fed into our classifier.We also define a custom preprocessing step in order to deal with language commonly used in interactions across many social media outlets. Our proposed system achieves a micro F1 score of 0.7340 on the test set and shows significant gains in performance compared to a system using dialogue level encoding.
U2 - 10.18653/v1/S19-2026
DO - 10.18653/v1/S19-2026
M3 - Article in proceedings
SP - 169
EP - 174
BT - Proceedings of the 13th International Workshop on Semantic Evaluation
PB - Association for Computational Linguistics
Y2 - 6 June 2019 through 7 June 2019
ER -
ID: 240420498