List your company
ResearchFormerly WizWand

Paraphrase Identification on Quora Question Pairs (test)

91.3AccuracyDRCN
79.13282.29185.4588.609Feb 13, 2017Dec 20, 2017Oct 26, 2018Sep 1, 2019Jul 7, 2020May 13, 2021Mar 20, 2022
Updated 8mo ago

Evaluation results

MethodLinks
DRCNEnsemble=true2018.0591.3
DRCNEnsemble=false2018.0590.15
DRCN2022.0390.1
DEIM2022.0390.1
DIINensemble=true2017.0989.84
DIINEnsemble=true2018.0589.84
SANsteps=52018.0489.4
SAN2019.0889.4
RE22022.0389.4
CSRAN2019.0889.2
RE22019.0889.2
DIINAttention Mechanism=Yes2018.0589.1
MwANAttention Mechanism=Yes2018.0589.1
Gong et al.2018.0489.1
DIIN2019.0889.1
MwAN2019.0889.1
DIINensemble=false2017.0989.06
DIINEnsemble=false2018.0589.06
pt-DECATT_char2017.0988.4
pt-DECATT charPre-training=Paralex, Type=character n-gram2017.0488.4
pt-DECATTcharAttention Mechanism=Yes2018.0588.4
pt-DecAttchar.cEnsemble=false2018.0588.4
Tomar et al.2018.0488.4
pt-DecAttn-charInput Granularity=character2019.0888.4
BIMPMAttention Mechanism=Yes2018.0588.2
SuBiLSTMBase Model=Siamese, Attention Mechanism=No2018.0588.2
BiMPM2019.0888.2
BIMPM2022.0388.2
BiMPM2017.0288.17
BiMPM2017.0988.17
BIMPM2017.0488.17
BiMPMEnsemble=false2018.0588.17
SuBiLSTM-TiedBase Model=Siamese, Attention Mechanism=No2018.0588.1
aESIM2018.1288.01
2-layer BiLSTMBase Model=Siamese, Attention Mechanism=No, Number of Layers=22018.0587.9
BiLSTMBase Model=Siamese, Attention Mechanism=No2018.0587.8
DECATT paralex-charPre-training=Paralex, Embeddings=character n-gram2017.0487.77
pt-DECATT_word2017.0987.54
pt-DECATT wordPre-training=Paralex, Type=word-based2017.0487.54
pt-DecAttn-wordInput Granularity=word2019.0887.5
ESIM2018.1286.98
DECATT charEmbeddings=character n-gram2017.0486.84
DECATT gloveEmbeddings=GloVe2017.0486.52
Capsule-BiGRU2022.0386.1
L.D.C.2017.0285.55
L.D.C2017.0985.55
L.D.C2017.0485.55
L.D.C.Ensemble=false2018.0585.55
L.D.C2018.1285.55
ESIM2022.0385.4
DECATT wordEmbeddings=None (word-based)2017.0485.27
FFNN charEmbeddings=character n-gram2017.0485.06
FFNN wordEmbeddings=uni-gram + tri-gram word2017.0484.35
Multi-Perspective-LSTM2017.0283.21
Multi-Perspective-LSTM2017.0983.21
Multi-Perspective-LSTM2017.0483.21
MP LSTMEnsemble=false2018.0583.21
Multi-Perspective-LSTM2018.1283.21
Siamese-LSTM2017.0282.58
Siamese-LSTM2017.0982.58
Siamese-LSTM2017.0482.58
Siamese-LSTMEnsemble=false2018.0582.58
Siamese-LSTM2018.1282.58
LSTM2022.0381.4
Multi-Perspective-CNN2017.0281.38
Multi-Perspective CNN2017.0981.38
Multi-Perspective CNN2017.0481.38
Multi perspective-CNN2018.1281.38
Siamese-CNN2017.0279.6
Siamese-CNN2017.0979.6
Siamese-CNN2017.0479.6
Siamese-CNN2018.1279.6