Table 7 Performance by Number of Positives for Training. This table shows results that are somewhat counter to expectation. For example, with BP there is a tendency for performance to drop with increasing numbers of positive example documents in the training set.

5

2

0.25

34

0.4067

128

0.2695

6–10

3

0.0833

25

0.3650

65

0.3875

11–15

9

0.4373

7

0.4528

22

0.3716

16–20

37

0.5645

4

0.4550

15

0.3306

21–25

39

0.544

3

0.4762

9

0.2588

26–30

31

0.5566

4

0.3687

4

0.3007

31–35

6

0.4663

3

0.5651

8

0.3566

36–40

7

0.5275

0

0

6

0.3579

41–45

10

0.4124

1

0.2009

5

0.3484

46–50

11

0.4276

1

0.2861

2

0.2553

51–75

18

0.3912

2

0.3430

12

0.3060

76–100

12

0.3936

1

0.2681

6

0.2726

101–125

5

0.4273

2

0.4089

0

0

126–150

4

0.4767

2

0.3226

0

0

151–last

20

0.3511

4

0.4586

1

0.2822