-
Notifications
You must be signed in to change notification settings - Fork 0
/
WSD.answers
96 lines (70 loc) · 1.39 KB
/
WSD.answers
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
1. Program is complete.
2. plant.wsd
3.
0 86.84
1 92.11
2 84.21
3 92.11
4 44.44
Average: 80.32
4.
Instance plant.1000004 (actual: living, predicted: factory)
Lots of token in data are generic tokens which appear in text belonging to both the senses. More training data with a
greater token diversity should help.
Instance plant.1000104 (actual: factory, predicted: living)
Many tokens in this text are things related to living beings (animal, fish, egg etc.) resulting in 'living' sense
probabilities winning. To correct this, we need to include more semantic understanding in our models (not possible in
Naive Bayes)
Instance plant.1000105 (actual: living, predicted: factory)
Number of tokens are less to get a meaningful discrimination of sense. Need more training data to get high quality
probabilities of features given sense.
--- bass.wsd ---
1. Program is complete.
2. bass.wsd
3.
0 72.73
1 95.45
2 90.91
3 100.0
4 94.74
Average: 90.65
--- crane.wsd ---
1. Program is complete.
2. crane.wsd
3.
0 68.42
1 73.68
2 73.68
3 89.47
4 84.21
Average: 77.89
--- motion.wsd ---
1. Program is complete.
2. motion.wsd
3.
0 63.41
1 78.05
2 90.24
3 90.24
4 89.19
Average: 82.09
--- palm.wsd ---
1. Program is complete.
2. palm.wsd
3.
0 75.61
1 75.61
2 68.29
3 80.49
4 75.68
Average: 75.12
--- tank-1.wsd ---
1. Program is complete.
2. tank.wsd
3.
0 68.29
1 56.1
2 82.93
3 82.93
4 45.95
Average: 67.66