Slot Filling on ATIS (test)
96.89F1 ScoreBi-model with a decoder
Evaluation Results
| Method | Links | |
|---|---|---|
| Bi-model with a decoderdecoder=with2018.12 | 96.89 | |
| Bi-model without a decoderdecoder=without2018.12 | 96.65 | |
| Attention BIRNN2018.12 | 95.98 | |
| Attention Encoder-Decoder NN2018.12 | 95.87 | |
| Ourstask_setting=multi-task2019.11 | 95.8 | |
| Attention Encoder-Decoder NN (with aligned inputs)Training mode=Independent, Aligned inputs=true2016.09 | 95.78 | |
| Attention BiRNNTraining mode=Independent2016.09 | 95.75 | |
| Encoder-labeler Deep LSTMTraining mode=Independent2016.09 | 95.66 | |
| Encoder-labeler Deep LSTM2018.12 | 95.66 | |
| AIN-1OEncoder=WORD-CHAR-BILSTM2020.09 | 95.59 | |
| AIN-F2OEncoder=WORD-CHAR-BILSTM2020.09 | 95.58 | |
| CRFEncoder=WORD-CHAR-BILSTM2020.09 | 95.52 | |
| SF-IDtask_setting=single-task2019.11 | 95.5 | |
| Joint GRU Model (W)2018.12 | 95.49 | |
| MaxEntEncoder=WORD-CHAR-BILSTM2020.09 | 95.47 | |
| Slotted-Gatedtask_setting=single-task2019.11 | 95.4 | |
| CNN CRF2018.12 | 95.35 | |
| RNN-EMTraining mode=Independent2016.09 | 95.25 | |
| RNN-EM2018.12 | 95.25 | |
| CAPSULE-NLUtask_setting=single-task2019.11 | 95.2 | |
| SPTIDtask_setting=single-task2019.11 | 95.1 | |
| GCDTtask_setting=single-task2019.11 | 95.1 | |
| Deep LSTMTraining mode=Independent2016.09 | 95.08 | |
| Hybrid RNNTraining mode=Independent2016.09 | 95.06 | |
| Hybrid RNN2018.12 | 95.06 | |
| Independent training RNN slot filling model2016.09 | 94.91 | |
| RNN with Label SamplingTraining mode=Independent2016.09 | 94.89 | |
| RNN with Label Sampling2018.12 | 94.89 | |
| CVTtask_setting=single-task2019.11 | 94.8 | |
| Joint model with recurrent slot label contextslot label context=recurrent2016.09 | 94.64 | |
| Joint model with recurrent slot label context2018.12 | 94.64 | |
| Joint model with local & recurrent slot label contextslot label context=local & recurrent2016.09 | 94.52 | |
| Joint model with recurrent intent + slot label contextintent context=recurrent, slot label context=yes2016.09 | 94.47 | |
| Joint model with recurrent intent and slot label context2018.12 | 94.47 | |
| Joint model with local & recurrent intent + slot label contextintent context=local & recurrent, slot label context=yes2016.09 | 94.45 | |
| CNN-CRFTraining mode=Independent2016.09 | 94.35 | |
| Joint Seqtask_setting=single-task2019.11 | 94.3 | |
| Joint model with local intent contextintent context=local2016.09 | 94.22 | |
| Attention-Basedtask_setting=single-task2019.11 | 94.2 | |
| Joint model with local & recurrent intent contextintent context=local & recurrent2016.09 | 94.18 | |
| Joint model with recurrent intent contextintent context=recurrent2016.09 | 94.16 | |
| Basic joint training model2016.09 | 94.15 | |
| Joint model with local slot label contextslot label context=local2016.09 | 94.14 | |
| Joint model with local intent + slot label contextintent context=local, slot label context=yes2016.09 | 94.13 | |
| RecNN+Viterbi2016.09 | 93.96 | |
| Recursive NN2018.12 | 93.96 | |
| RecNN2016.09 | 93.22 | |
| AIN-F2OEncoder=WORD-CNN2020.09 | 92.26 | |
| CRFEncoder=WORD-CNN2020.09 | 92.24 | |
| AIN-1OEncoder=WORD-CNN2020.09 | 92.18 | |
| MaxEntEncoder=WORD-CNN2020.09 | 91.07 | |
| AIN-F2OEncoder=WORD ONLY2020.09 | 88.86 | |
| AIN-1OEncoder=WORD ONLY2020.09 | 88.76 | |
| CRFEncoder=WORD ONLY2020.09 | 76.82 | |
| MaxEntEncoder=WORD ONLY2020.09 | 62.93 |