Baseline Disambiguation Using First Sense

Prepositin

Number of Senses

 Instances

Recall

Fine-Grained Precision

Coarse-Grained Precision

about

6

364

1.000

0.885

0.907

above

9

23

1.000

0.043

0.217

across

3

153

1.000

0.961

0.961

after

11

53

1.000

0.566

0.566

against

10

92

1.000

0.446

0.522

along

4

174

1.000

0.954

0.954

among

4

50

1.000

0.300

0.300

around

6

155

1.000

0.155

0.155

as

2

84

1.000

1.000

1.000

at

12

367

1.000

0.425

0.425

before

4

20

1.000

0.450

0.450

behind

9

68

1.000

0.662

0.662

beneath

4

28

1.000

0.571

0.857

beside

6

29

1.000

1.000

1.000

between

3

102

1.000

0.422

0.422

by

9

249

1.000

0.000

0.554

down

9

153

1.000

0.438

0.993

during

22

39

1.000

0.615

1.000

for

5

479

1.000

0.035

0.035

from

2

581

1.000

0.279

0.282

in

15

689

1.000

0.373

0.492

inside

16

38

1.000

0.368

1.000

into

15

297

1.000

0.293

0.293

like

5

125

1.000

0.120

0.984

of

10

1480

1.000

0.000

0.315

off

7

76

1.000

0.171

0.171

on

5

444

1.000

0.216

0.468

onto

20

58

1.000

0.879

0.879

over

7

98

1.000

0.010

0.316

round

25

82

1.000

0.037

0.073

through

3

208

1.000

0.322

0.885

to

17

572

1.000

0.322

0.327

towards

8

102

1.000

0.873

0.873

with

16

579

1.000

0.149

0.149

Total

17

8111

1.000

0.291

0.441



Most Frequent Senses (Based on Occurrences in Training Data)

Preposition

Number of Senses

Sense

Number of Occurrences

Number of Instances

Frequency of Occurrence

about

6

1(1)

637.0

710

0.90

above

9

4(2)

27.0

48

0.56

across

3

1(1)

307.0

319

0.96

after

11

1(1)

37.0

103

0.36

against

10

10(4)

84.0

195

0.43

along

4

1(1)

348.0

365

0.95

among

4

1(1)

41.0

100

0.41

around

6

3(2)

165.0

335

0.49

as

2

1(1)

174.0

174

1.00

at

12

1(1)

277.0

715

0.39

before

4

1(1)

27.0

47

0.57

behind

9

1(1)

89.0

138

0.64

beneath

4

1(1)

37.0

57

0.65

beside

6

1(1)

62.0

62

1.00

between

3

1(1)

86.0

211

0.41

by

9

3(1b)

217.0

510

0.43

down

9

1(1)

163.0

332

0.49

during

22

2(1a)

41.0

81

0.51

for

5

5(4)

267.0

951

0.28

from

2

1(1)

330.0

1206

0.27

in

15

1(1)

520.0

1397

0.37

inside

16

2(1a)

27.0

67

0.40

into

15

3(3)

270.5

604

0.45

like

5

2(1a)

202.0

266

0.76

of

10

3(1b)

642.0

3004

0.21

off

7

4(3)

118.0

161

0.73

on

5

8(3)

173.0

872

0.20

onto

20

1(1)

104.0

117

0.89

over

7

16(7)

78.0

200

0.39

round

25

8(4)

59.0

181

0.33

through

3

3(1b)

165.5

441

0.38

to

17

1(1)

403.0

1183

0.34

towards

8

1(1)

187.0

214

0.87

with

16

5(3a)

289.0

1187

0.24



Baseline Disambiguation Using Most Frequent Training Sense

Prepostition

Number of Senses

 Instances

Recall

Fine-Grained Precision

Coarse-Grained Precision

about

6

364

1.000

0.885

0.907

above

9

23

1.000

0.609

0.783

across

3

153

1.000

0.961

0.961

after

11

53

1.000

0.566

0.566

against

10

92

1.000

0.435

0.435

along

4

174

1.000

0.954

0.954

among

4

50

1.000

0.300

0.300

around

6

155

1.000

0.452

0.452

as

2

84

1.000

1.000

1.000

at

12

367

1.000

0.425

0.425

before

4

20

1.000

0.450

0.450

behind

9

68

1.000

0.662

0.662

beneath

4

28

1.000

0.571

0.857

beside

6

29

1.000

1.000

1.000

between

3

102

1.000

0.422

0.422

by

9

249

1.000

0.373

0.554

down

9

153

1.000

0.438

0.993

during

22

39

1.000

0.385

1.000

for

5

479

1.000

0.238

0.238

from

2

581

1.000

0.279

0.282

in

15

689

1.000

0.373

0.492

inside

16

38

1.000

0.526

1.000

into

15

297

1.000

0.451

0.438

like

5

125

1.000

0.768

0.984

of

10

1480

1.000

0.268

0.315

off

7

76

1.000

0.763

0.763

on

5

444

1.000

0.205

0.282

onto

20

58

1.000

0.879

0.879

over

7

98

1.000

0.327

0.327

round

25

82

1.000

0.378

0.378

through

3

208

1.000

0.495

0.885

to

17

572

1.000

0.322

0.327

towards

8

102

1.000

0.873

0.873

with

16

579

1.000

0.249

0.394

Total

17

8111

1.000

0.410

0.480