Multi-Lingual Dialogue Act Recognition with Deep Learning Methods

Jiří Martínek (1) , Pavel Kral (1) , Ladislav Lenc (1) , Christophe Cerisara (2)

1 University of West Bohemia [Plzeň ]
2 SYNALP - Natural Language Processing : representations, inference and semantics

This paper deals with multilingual dialogue act (DA) recognition. The proposed approaches are based on deep neural networks and use word2vec embeddings for word representation. Two multilingual models are proposed for this task. The first approach uses one general model trained on the embeddings from all available languages. The second method trains the model on a single pivot language and a linear transformation method is used to project other languages onto the pivot language. The popular convolutional neural network and LSTM architectures with different setups are used as classifiers. To the best of our knowledge this is the first attempt at multilingual DA recognition using neural networks. The multilingual models are validated experimentally on two languages from the Verbmobil corpus.

Mots clés: Index Terms: CNN; deep learning; dialogue act; LSTM; multi- linguality; word embeddings; word2vec

Domaines: Traitement du texte et du document

Jiří Martínek, Pavel Kral, Ladislav Lenc, Christophe Cerisara. Multi-Lingual Dialogue Act Recognition with Deep Learning Methods. Interspeech 2019, Sep 2019, Graz, Austria. ⟨10.21437/Interspeech.2019-1691⟩. ⟨hal-02319818⟩