List your company

Dialog act prediction on MRDA (test)

92.2AccuracyDAH-CRF + LDAutt
57.77666.71375.6584.587Mar 12, 2016Feb 9, 2017Jan 10, 2018Dec 11, 2018Nov 10, 2019Oct 10, 2020Sep 10, 2021

Evaluation results

MethodLinks
DAH-CRF + LDAutttopic labels=automatically acquired utterance-level via LDA2018.1092.2
Li et al. (2018a)2020.0292.2
CRF-ASN2017.1191.7
Chen et al. (2018)2019.0491.7
Chen et al. (2018)2020.0291.7
seq2seqBESTencoder=HGRU, decoder=hard guided attention, Btrain=5, Binf=12020.0291.6
Seq2Seq2021.0991.6
CRF-ASNreported_from=prior publication2018.1091.4
CRF-ASN2021.0991.4
Oursspeaker turn embeddings=true, topic-aware embeddings=false2021.0991.4
DAH-CRF + LDAconvtopic labels=automatically acquired conversation-level via LDA2018.1091.2
Our Method2019.0491.1
SelfAtt-CRFreported_from=prior publication2018.1091.1
Raheja and Tetreault (2019)2020.0291.1
SelfAtt-CRF2021.0991.1
Hierarchical Bi-LSTM-CRF2017.0990.9
Bi-LSTM-CRF2017.1190.9
Kumar et al. (2018)2019.0490.9
Bi-LSTM-CRFreported_from=prior publication2018.1090.9
Kumar et al. (2018)2020.0290.9
Bi-LSTM-CRF2021.0990.9
Ours-Speakerspeaker turn embeddings=false, topic-aware embeddings=false2021.0990.7
DRLM-Cond2018.1088.4
DRLM-Cond2021.0988.4
LSTM-Softmax2017.0986.8
LSTM-Softmax2017.1186.8
Khanpour et al. (2016)2019.0486.8
SGNN2021.0986.7
CNNselection=highest validation accuracy run2016.0384.6
CNN2017.0984.6
CNN2017.1184.6
Lee and Dernoncourt (2016)2019.0484.6
LSTMselection=highest validation accuracy run2016.0384.3
Ortega and Vu (2017)2019.0484.3
CRF2017.1183.9
Naive Bayesfeatures=transcribed words2016.0382
Naiive Bayes2017.0982
SVM2017.1182
Graphical modelfeatures=transcribed words and previous predicted dialog acts2016.0381.3
JAS2018.1081.3
TF-IDF GloVe2019.0478.7
Majority class2016.0359.1