DAFNA-EA/ParametersNamesDefaultValues.txt at master · qcri/DAFNA-EA · GitHub

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
General Parameters:
-------------------
- cosineSimDiffStoppingCriteria: The difference beteween the Sources trustworthiness cosine similarity between 2 iterations must be  less than this value for a convergence sign.
Should be very small double value, default 0.001.(double)

- startingTrust: The value used for initializing all sources trustworthiness. Default Value 0.8.(double)

- startingErrorFactor: The value used for initializing all values error factor(if any- Only used for 3-Estimates). default value 0.4.(double)

- startingConfidence: The value used for initializing all valuesConfidence (if any- Only used for Cosine). default value 1.0.(double)

Cosine:
-------
- dampeningFactorCosine: The used Dampeneing Factor.
from zero to one, default value 0.2. (double value)

2-Estimates:
------------
-normalizationWeight: The Normalization weight.
From zero to one, default Value 0.5. (double value)

3-Estimates:
------------
-normalizationWeight: The Normalization weight.
 From zero to one, default Value 0.5. (double value)

LTM (Latent Truth Model):
----
- (bita1, bita0)  default value (0.5, 0.5) (double value,  double value)
bita1: prior true count.
bita0: prior false count.
- (alpha01, alpha00)  default value (0.9, 0.1) (double value,  double value)
alpha01: Prior false positive count.
alpha00: prior true negative count.

- (alpha11, alpha10)  default value (0.9, 0.1) (double value,  double value)
alpha11: Prior true positive count.
alpha10: prior false negative count.

-(K, B, T): The collapsed Gipps Sampling parameters.
- K: Number of Iterations default value 500 (integer)
- B: Burn-in period default value 100 (integer)
- T: Thinning default value 9 (ineger)


Source Dependency-Aware Models: Depen, Accu, AccuSim, AccuNoDep
----------------------------------------------------------------
(Depen:
    boolean considerSimilarity = false;
    boolean considerSourcesAccuracy = false;
    boolean computeNormalDependency = true;
)

(Accu:
    boolean considerSimilarity = false;
    boolean considerSourcesAccuracy = true;
    boolean computeNormalDependency = true;
)

(AccuSim:
    boolean considerSimilarity = true;
    boolean considerSourcesAccuracy = true;
    boolean computeNormalDependency = true;
)

(AccuNoDep:
    boolean considerSimilarity = false;
    boolean considerSourcesAccuracy = true;
    boolean computeNormalDependency = false;
)

- alpha: Apriori probability that 2 sources are dependent, from zero to 0.5. Default value = 0.2.
- c: The probability that a value provided by a copier is copied. From zero to one, (double)
- n: Number of false value per data item. Default value 100. (Integer). should be zero


TruthFinder
-----------
- similarityConstant: The similarity Constant.
From zero to One, default value = 0.5(double)

- dampingFactor: The dampning Factor, from zero to one, default value 0.1.(double)

Simple LCA AND Guess LCA
-------------------------
- Bita1: Overall prior truth probability. From zero to one, default value 0.5(double)

MLE
---
- Bita1: Overall prior truth probability. From zero to one, default value 0.5(double)
- r: The probabilty that a source provide a value for all data items. From zero to one, default value 0.5.(double)