Chinese Task . | Document Genre . | Question Type . | Answer Type . | Question Size . | English Counterpart . |
---|---|---|---|---|---|
Question Answering | |||||
QS (Cheng et al., 2016) | N/A | free-form | multiple-choice | 0.6K | ARC (Clark et al., 2016) |
MCQA (Guo et al., 2017a) | N/A | free-form | multiple-choice | 14.4K | ARC (Clark et al., 2016) |
MedQA (Zhang et al., 2018b) | N/A | free-form | multiple-choice | 235.2K | ARC (Clark et al., 2016) |
GeoSQA (Huang et al., 2019) | N/A | free-form | multiple-choice | 4.1K | DD (Lally et al., 2017) |
Machine Reading Comprehension | |||||
PD (Cui et al., 2016) | news | cloze | extractive | 876.7K | CNN/Daily (Hermann et al., 2015) |
CFT (Cui et al., 2016) | books | cloze | extractive | 3.6K | CBT (Hill et al., 2016) |
CMRC 2018 (Cui et al., 2018b) | Wiki | free-form | extractive | 19.1K | SQuAD (Rajpurkar et al., 2016) |
DuReader (He et al., 2017) | web | free-form | abstractive | ≈ 200K | MS MARCO (Nguyen et al., 2016) |
ChID (Zheng et al., 2019) | mixed-genre | cloze | multiple-choice | 728.7K | CLOTH (Xie et al., 2018) |
CM3(this work) | mixed-genre | free-form | multiple-choice | 10.0K | RACE (Lai et al., 2017) |
CD3(this work) | dialogue | free-form | multiple-choice | 9.6K | DREAM (Sun et al., 2019a) |
Chinese Task . | Document Genre . | Question Type . | Answer Type . | Question Size . | English Counterpart . |
---|---|---|---|---|---|
Question Answering | |||||
QS (Cheng et al., 2016) | N/A | free-form | multiple-choice | 0.6K | ARC (Clark et al., 2016) |
MCQA (Guo et al., 2017a) | N/A | free-form | multiple-choice | 14.4K | ARC (Clark et al., 2016) |
MedQA (Zhang et al., 2018b) | N/A | free-form | multiple-choice | 235.2K | ARC (Clark et al., 2016) |
GeoSQA (Huang et al., 2019) | N/A | free-form | multiple-choice | 4.1K | DD (Lally et al., 2017) |
Machine Reading Comprehension | |||||
PD (Cui et al., 2016) | news | cloze | extractive | 876.7K | CNN/Daily (Hermann et al., 2015) |
CFT (Cui et al., 2016) | books | cloze | extractive | 3.6K | CBT (Hill et al., 2016) |
CMRC 2018 (Cui et al., 2018b) | Wiki | free-form | extractive | 19.1K | SQuAD (Rajpurkar et al., 2016) |
DuReader (He et al., 2017) | web | free-form | abstractive | ≈ 200K | MS MARCO (Nguyen et al., 2016) |
ChID (Zheng et al., 2019) | mixed-genre | cloze | multiple-choice | 728.7K | CLOTH (Xie et al., 2018) |
CM3(this work) | mixed-genre | free-form | multiple-choice | 10.0K | RACE (Lai et al., 2017) |
CD3(this work) | dialogue | free-form | multiple-choice | 9.6K | DREAM (Sun et al., 2019a) |