Class: BayesMotel::Corpus

Inherits:
Object
  • Object
show all
Defined in:
lib/bayes_motel/corpus.rb

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(name) ⇒ Corpus

Returns a new instance of Corpus.



7
8
9
10
11
 
# File 'lib/bayes_motel/corpus.rb', line 7

def initialize(name)
  @name = name
  @total_count = 0
  @data = {}
end

Instance Attribute Details

#data ⇒ Object (readonly)

Returns the value of attribute data.



5
6
7
 
# File 'lib/bayes_motel/corpus.rb', line 5

def data
  @data
end

#name ⇒ Object (readonly)

Returns the value of attribute name.



3
4
5
 
# File 'lib/bayes_motel/corpus.rb', line 3

def name
  @name
end

#total_count ⇒ Object (readonly)

Returns the value of attribute total_count.



4
5
6
 
# File 'lib/bayes_motel/corpus.rb', line 4

def total_count
  @total_count
end

Instance Method Details

#classify(doc) ⇒ Object

The default classification algorithm just picks the category with the highest score.



24
25
26
27
28
29
30
31
 
# File 'lib/bayes_motel/corpus.rb', line 24

def classify(doc)
  results = score(doc)
  max = [:none, 0]
  results.each_pair do |(k, v)|
    max = [k, v] if v > max[1]
  end
  max
end

#cleanup ⇒ Object 



33
34
35
36
37
 
# File 'lib/bayes_motel/corpus.rb', line 33

def cleanup
  @data.each_pair do |k, v|
    clean(@data, k, v)
  end
end

#score(doc) ⇒ Object 



18
19
20
 
# File 'lib/bayes_motel/corpus.rb', line 18

def score(doc)
  _score(doc)
end

#train(doc, category) ⇒ Object 



13
14
15
16
 
# File 'lib/bayes_motel/corpus.rb', line 13

def train(doc, category)
  @total_count += 1
  _training(doc, category)
end