Inter-Turn Relation Prediction on CMV QR data benchmark leaderboard · SOTA2 Research

Inter-Turn Relation Prediction on CMV QR data

18.9PrecisionQR Context Fine-Tuned BERT + RST Features + Extractive Summarizer + Source ≠ Target Constraint

Evaluation Results

Method
QR Context Fine-Tuned BERT + RST Features + Extractive Summarizer + Source ≠ Target Constraint2020.04	18.9	79	30.5
QR Context Fine-Tuned BERT + RST Features + Extractive Summarizer + Source ≠ Target Constraint2020.04	17.5	74	28.3
QR Context Fine-Tuned BERT + RST Features + Extractive Summarizer2020.04	16	79.4	26.8
QR Context Fine-Tuned BERT + RST Features + Extractive Summarizer2020.04	14.5	75.6	24.3
QR Context Fine-Tuned BERT + RST Features2020.04	12.2	75.5	19.1
QR Context Fine-Tuned BERT2020.04	11	75.3	19.1
QR Context Fine-Tuned BERT + RST Features2020.04	11	79	21.2
QR Context Fine-Tuned BERT2020.04	10	72.5	17.6
BERT2020.04	8.8	76	15.8
BERT2020.04	7.9	70	14.1
Morio and Fujita (2018)2020.04	7.6	40	12.7
Menini et al. (2018) + RST Features2020.04	6.4	83	11.8
Menini et al. (2018)2020.04	5.9	82	11
RST Features2020.04	5.1	80	9.6
All relations2020.04	5	100	9
Menini et al. (2018) + RST Features2020.04	4.9	80	9.3
Menini et al. (2018)2020.04	4.8	80	9
RST Features2020.04	3.8	77	7.2