![[Previous]](http://snow.ucsd.edu/~ezra/msc/GIF/claspb1w.gif){kind=small}
![[Next]](http://snow.ucsd.edu/~ezra/msc/GIF/claspb2w.gif){kind=small}
![[Chapter]](http://snow.ucsd.edu/~ezra/msc/GIF/claspb3w.gif){kind=small}
![[Contents]](http://snow.ucsd.edu/~ezra/msc/GIF/claspb4w.gif){kind=small}
![[Search]](http://snow.ucsd.edu/~ezra/msc/GIF/claspb5.gif){kind=small}
![[Feedback]](http://snow.ucsd.edu/~ezra/msc/GIF/claspb6w.gif){kind=small}
|
|
A.5.2.2 Increasing the Size of the Training Corpus
#100 #500 #1000 #2000 #3000 #4000
Total: 59965 59965 59965 59965 59965 59965
Indicative: 30700 30700 30700 30700 30700 30700
On: 5061 1685 573 900 351 353
Mis: 16% 5% 1% 2% 1% 1%
Off: 3097 1941 1622 554 1068 811
Spu: 10% 6% 5% 1% 3% 2%
Bad: 26% 11% 7% 4% 4% 3%
Gud: 86% 93% 96% 97% 97% 98%
Imperative: 2199 2199 2199 2199 2199 2199
On: 1331 1144 803 518 475 371
Mis: 60% 52% 36% 23% 21% 16%
Off: 1528 485 323 229 219 266
Spu: 2% 0% 0% 0% 0% 0%
Bad: 130% 74% 51% 33% 31% 28%
Gud: 95% 97% 98% 98% 98% 98%
Interrogative: 19943 19943 19943 19943 19943 19943
On: 3710 2145 1682 692 1015 963
Mis: 18% 10% 8% 3% 5% 4%
Off: 4710 2067 845 1214 609 579
Spu: 11% 5% 2% 3% 1% 1%
Bad: 42% 21% 12% 9% 8% 7%
Gud: 85% 92% 95% 96% 97% 97%
ID1: 35597 35597 35597 35597 35597 35597
On: 4888 2232 2294 1484 1653 810
Mis: 13% 6% 6% 4% 4% 2%
Off: 9391 4173 3020 2246 841 821
Spu: 38% 17% 12% 9% 3% 3%
Bad: 40% 17% 14% 10% 7% 4%
Gud: 76% 89% 91% 93% 95% 97%
ID2: 17978 17978 17978 17978 17978 17978
On: 8340 3913 2811 1437 814 698
Mis: 46% 21% 15% 7% 4% 3%
Off: 5110 1799 1671 1465 1431 550
Spu: 12% 4% 3% 3% 3% 1%
Bad: 74% 31% 24% 16% 12% 6%
Gud: 77% 90% 92% 95% 96% 97%
ID3: 8 8 8 8 8 8
On: 8 8 8 8 8 8
Mis: 100% 100% 100% 100% 100% 100%
Off: 0 0 1 0 0 0
Spu: 0% 0% 0% 0% 0% 0%
Bad: 100% 100% 112% 100% 100% 100%
Gud: 99% 99% 99% 99% 99% 99%
Status: 40437 40437 40437 40437 40437 40437
On: 3167 1444 2086 1168 972 539
Mis: 7% 3% 5% 2% 2% 1%
Off: 7534 3244 2009 1296 880 597
Spu: 38% 16% 10% 6% 4% 3%
Bad: 26% 11% 10% 6% 4% 2%
Gud: 82% 92% 93% 95% 96% 98%
Declarative: 20154 20154 20154 20154 20154 20154
On: 5140 1996 1502 2898 1708 1344
Mis: 25% 9% 7% 14% 8% 6%
Off: 7530 4451 4016 1606 1645 1325
Spu: 18% 11% 10% 4% 4% 3%
Bad: 62% 31% 27% 22% 16% 13%
Gud: 78% 89% 90% 92% 94% 95%
Order: 3560 3560 3560 3560 3560 3560
On: 2582 1706 1440 1438 1099 751
Mis: 72% 47% 40% 40% 30% 21%
Off: 2596 1346 812 485 497 563
Spu: 4% 2% 1% 0% 0% 0%
Bad: 145% 85% 63% 54% 44% 36%
Gud: 91% 94% 96% 96% 97% 97%
WH-Question: 13511 13511 13511 13511 13511 13511
On: 4559 1550 1820 830 722 383
Mis: 33% 11% 13% 6% 5% 2%
Off: 3907 1610 538 587 459 516
Spu: 8% 3% 1% 1% 0% 1%
Bad: 62% 23% 17% 10% 8% 6%
Gud: 85% 94% 96% 97% 98% 98%
YN-Question: 5541 5541 5541 5541 5541 5541
On: 3987 2585 2212 2044 1033 1101
Mis: 71% 46% 39% 36% 18% 19%
Off: 3354 1491 1455 1305 1281 1109
Spu: 6% 2% 2% 2% 2% 2%
Bad: 132% 73% 66% 60% 41% 39%
Gud: 87% 93% 93% 94% 96% 96%
Relative: 3731 3731 3731 3731 3731 3731
On: 2705 1322 1099 659 533 265
Mis: 72% 35% 29% 17% 14% 7%
Off: 2381 686 643 663 424 223
Spu: 4% 1% 1% 1% 0% 0%
Bad: 136% 53% 46% 35% 25% 13%
Gud: 91% 96% 97% 97% 98% 99%
Connector: 4288 4288 4288 4288 4288 4288
On: 3334 1547 1084 488 544 496
Mis: 77% 36% 25% 11% 12% 11%
Off: 1745 817 546 1400 830 623
Spu: 3% 1% 0% 2% 1% 1%
Bad: 118% 55% 38% 44% 32% 26%
Gud: 91% 96% 97% 96% 97% 98%
Complement: 1445 1445 1445 1445 1445 1445
On: 1445 961 658 406 381 381
Mis: 100% 66% 45% 28% 26% 26%
Off: 0 745 627 916 606 497
Spu: 0% 1% 1% 1% 1% 0%
Bad: 100% 118% 88% 91% 68% 60%
Gud: 97% 97% 97% 97% 98% 98%
Infinity: 3374 3374 3374 3374 3374 3374
On: 2833 1260 617 357 365 210
Mis: 83% 37% 18% 10% 10% 6%
Off: 1096 838 653 484 215 146
Spu: 1% 1% 1% 0% 0% 0%
Bad: 116% 62% 37% 24% 17% 10%
Gud: 93% 96% 97% 98% 99% 99%
Voice: 39166 39166 39166 39166 39166 39166
On: 9735 6591 6960 2937 2016 6454
Mis: 24% 16% 17% 7% 5% 16%
Off: 6997 6461 6239 7616 6795 5562
Spu: 33% 31% 29% 36% 32% 26%
Bad: 42% 33% 33% 26% 22% 30%
Gud: 72% 78% 77% 82% 85% 79%
Polarity: 27213 27213 27213 27213 27213 27213
On: 13339 11151 4276 11013 12219 9481
Mis: 49% 40% 15% 40% 44% 34%
Off: 10133 7563 13260 8453 6660 8869
Spu: 30% 23% 40% 25% 20% 27%
Bad: 86% 68% 64% 71% 69% 67%
Gud: 60% 68% 70% 67% 68% 69%
CST: 88 94 95 95 96 97