August 17, 2005
Generic Soft Pattern Models for
Definitional QA
24/28
Sensitivity
to the Amount of Training
Data
•PHMM requires more training data
to improve
0.2553 (+3.03%)
0.2510 (+1.29%)
0.2478
Bigram R3E
0.2303 (+3.32%)
0.2269 (+1.76%)
0.2229
Bigram R3A
0.2496
(+8.00%)
0.2402
(+3.93%)
0.2311
PHMM R3E
0.2234
(+5.85%)
0.2179
(+3.24%)
0.2110
PHMM R3A
1
1/2
1/3
Training Data size
(fraction of whole training corpus)
7.22%
2.28%