Answer Ranking on TREC QA (train)
0.7495MAPaNMM-1
Evaluation results
| Method | Links | ||
|---|---|---|---|
| aNMM-1combining additional features=true2018.01 | 0.7495 | 0.8109 | |
| aNMMtraining_setting=best setting trained on TRAIN-ALL2018.01 | 0.7495 | 0.8109 | |
| aNMM-2combining additional features=true2018.01 | 0.7484 | 0.8013 | |
| Severyn et al. (2015)combining additional features=true2018.01 | 0.7459 | 0.8078 | |
| Severyn et al. (2015)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.7459 | 0.8078 | |
| aNMM-1combining additional features=true2018.01 | 0.7417 | 0.8102 | |
| aNMM-2additional_features=none, method_type=deep learning2018.01 | 0.7407 | 0.7969 | |
| aNMM-2Additional Features=None2018.01 | 0.7407 | 0.7969 | |
| aNMM-1additional_features=none, method_type=deep learning2018.01 | 0.7385 | 0.7995 | |
| aNMM-1Additional Features=None2018.01 | 0.7385 | 0.7995 | |
| aNMM-1Additional Features=None2018.01 | 0.7334 | 0.802 | |
| Severyn et al. (2015)combining additional features=true2018.01 | 0.7329 | 0.7962 | |
| aNMM-2combining additional features=true2018.01 | 0.7306 | 0.7968 | |
| aNMM-2Additional Features=None2018.01 | 0.7191 | 0.7974 | |
| Wang et al. (2015)combining additional features=true2018.01 | 0.7134 | 0.7913 | |
| Wang et al. (2015)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.7134 | 0.7913 | |
| Yu et al. (2014)combining additional features=true2018.01 | 0.7113 | 0.7846 | |
| Yu et al. (2014)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.7113 | 0.7846 | |
| Yih et al. (2013)additional_features=none, method_type=feature engineering2018.01 | 0.7092 | 0.77 | |
| Yih et al. (2013)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.7092 | 0.77 | |
| Yu et al. (2014)combining additional features=true2018.01 | 0.7058 | 0.78 | |
| Severyn et al. (2013)additional_features=none, method_type=feature engineering2018.01 | 0.6781 | 0.7358 | |
| Severyn et al. (2013)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.6781 | 0.7358 | |
| Severyn et al. (2015)Additional Features=None2018.01 | 0.6709 | 0.728 | |
| Yao et al. (2013)additional_features=none, method_type=feature engineering2018.01 | 0.6307 | 0.7477 | |
| Yao et al. (2013)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.6307 | 0.7477 | |
| Severyn et al. (2015)Additional Features=None2018.01 | 0.6258 | 0.6591 | |
| Heilman and Smith (2010)additional_features=none, method_type=feature engineering2018.01 | 0.6091 | 0.6917 | |
| Heilman and Smith (2010)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.6091 | 0.6917 | |
| Wang et al. (2007)additional_features=none, method_type=feature engineering2018.01 | 0.6029 | 0.6852 | |
| Wang et al. (2007)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.6029 | 0.6852 | |
| Wang and Manning (2010)additional_features=none, method_type=feature engineering2018.01 | 0.5951 | 0.6951 | |
| Wang and Manning (2010)training_setting=best setting trained on TRAIN-ALL2018.01 | 0.5951 | 0.6951 | |
| Wang et al. (2015)Additional Features=None2018.01 | 0.5928 | 0.6721 | |
| Yu et al. (2014)Additional Features=None2018.01 | 0.5693 | 0.6613 | |
| Yu et al. (2014)Additional Features=None2018.01 | 0.5476 | 0.6437 |