Class: StuffClassifier::Base

Inherits:
Object
  • Object
show all
Extended by:
Storage::ActAsStorable
Defined in:
lib/stuff-classifier/base.rb

Direct Known Subclasses

Bayes, TfIdf

Class Attribute Summary collapse

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Methods included from Storage::ActAsStorable

storable, to_store

Constructor Details

#initialize(name, opts = {}) ⇒ Base

opts : language stemming : true | false weight assumed_prob storage purge_state ?



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
 
# File 'lib/stuff-classifier/base.rb', line 27

def initialize(name, opts={})
  @version = StuffClassifier::VERSION
  
  @name = name

  # This values are nil or are loaded from storage
  @word_list = {}
  @category_list = {}
  @training_count=0

  # storage
  purge_state = opts[:purge_state]
  @storage = opts[:storage] || StuffClassifier::Base.storage
  unless purge_state
    @storage.load_state(self)
  else
    @storage.purge_state(self)
  end

  # This value can be set during initialization or overrided after load_state
  @thresholds = opts[:thresholds] || {}
  @min_prob = opts[:min_prob] || 0.0
  

  @ignore_words = nil
  @tokenizer = StuffClassifier::Tokenizer.new(opts)
  
end

Class Attribute Details

.storage ⇒ Object 



175
176
177
178
 
# File 'lib/stuff-classifier/base.rb', line 175

def storage
  @storage = StuffClassifier::InMemoryStorage.new unless defined? @storage
  @storage
end

Instance Attribute Details

#category_list ⇒ Object (readonly)

Returns the value of attribute category_list.



7
8
9
 
# File 'lib/stuff-classifier/base.rb', line 7

def category_list
  @category_list
end

#language ⇒ Object

Returns the value of attribute language.



11
12
13
 
# File 'lib/stuff-classifier/base.rb', line 11

def language
  @language
end

#min_prob ⇒ Object

Returns the value of attribute min_prob.



14
15
16
 
# File 'lib/stuff-classifier/base.rb', line 14

def min_prob
  @min_prob
end

#name ⇒ Object (readonly)

Returns the value of attribute name.



5
6
7
 
# File 'lib/stuff-classifier/base.rb', line 5

def name
  @name
end

#thresholds ⇒ Object

Returns the value of attribute thresholds.



13
14
15
 
# File 'lib/stuff-classifier/base.rb', line 13

def thresholds
  @thresholds
end

#tokenizer ⇒ Object

Returns the value of attribute tokenizer.



10
11
12
 
# File 'lib/stuff-classifier/base.rb', line 10

def tokenizer
  @tokenizer
end

#training_count ⇒ Object (readonly)

Returns the value of attribute training_count.



8
9
10
 
# File 'lib/stuff-classifier/base.rb', line 8

def training_count
  @training_count
end

#word_list ⇒ Object (readonly)

Returns the value of attribute word_list.



6
7
8
 
# File 'lib/stuff-classifier/base.rb', line 6

def word_list
  @word_list
end

Class Method Details

.open(name) ⇒ Object 



180
181
182
183
184
185
186
187
188
 
# File 'lib/stuff-classifier/base.rb', line 180

def open(name)
  inst = self.new(name)
  if block_given?
    yield inst
    inst.save_state
  else
    inst
  end
end

Instance Method Details

#cat_count(category) ⇒ Object

return the number of training document for a category



108
109
110
 
# File 'lib/stuff-classifier/base.rb', line 108

def cat_count(category)
  @category_list[category][:_count] ? @category_list[category][:_count].to_f : 0.0
end

#categories ⇒ Object

return categories list



124
125
126
 
# File 'lib/stuff-classifier/base.rb', line 124

def categories
  @category_list.keys
end

#categories_with_word_count(word) ⇒ Object

return the number of time categories in wich a word appear



113
114
115
116
 
# File 'lib/stuff-classifier/base.rb', line 113

def categories_with_word_count(word)
  return 0 unless @word_list[word] && @word_list[word][:categories]
  @word_list[word][:categories].length 
end

#classify(text, default = nil) ⇒ Object

classify a text



135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
 
# File 'lib/stuff-classifier/base.rb', line 135

def classify(text, default=nil)
  # Find the category with the highest probability
  max_prob = @min_prob
  best = nil

  scores = cat_scores(text)
  scores.each do |score|
    cat, prob = score
    if prob > max_prob
      max_prob = prob
      best = cat
    end
  end
  
  # Return the default category in case the threshold condition was
  # not met. For example, if the threshold for :spam is 1.2
  #
  #    :spam => 0.73, :ham => 0.40  (OK)
  #    :spam => 0.80, :ham => 0.70  (Fail, :ham is too close)

  return default unless best

  threshold = @thresholds[best] || 1.0

  scores.each do |score|
    cat, prob = score
    next if cat == best
    return default if prob * threshold > max_prob
  end

  return best    
end

#incr_cat(category) ⇒ Object 



74
75
76
77
78
79
80
81
82
 
# File 'lib/stuff-classifier/base.rb', line 74

def incr_cat(category)
  @category_list[category] ||= {}
  @category_list[category][:_count] ||= 0
  @category_list[category][:_count] += 1

  @training_count ||= 0
  @training_count += 1 

end

#incr_word(word, category) ⇒ Object 



56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
 
# File 'lib/stuff-classifier/base.rb', line 56

def incr_word(word, category)
  @word_list[word] ||= {}

  @word_list[word][:categories] ||= {}
  @word_list[word][:categories][category] ||= 0
  @word_list[word][:categories][category] += 1

  @word_list[word][:_total_word] ||= 0
  @word_list[word][:_total_word] += 1


  # words count by categroy
  @category_list[category] ||= {}
  @category_list[category][:_total_word] ||= 0
  @category_list[category][:_total_word] += 1

end

#save_state ⇒ Object 



168
169
170
 
# File 'lib/stuff-classifier/base.rb', line 168

def save_state
  @storage.save_state(self)
end

#total_cat_count ⇒ Object

return the number of training item



103
104
105
 
# File 'lib/stuff-classifier/base.rb', line 103

def total_cat_count
  @training_count
end

#total_categories ⇒ Object

return the number of categories



119
120
121
 
# File 'lib/stuff-classifier/base.rb', line 119

def total_categories
  categories.length
end

#total_word_count(word) ⇒ Object

return the number of times the word appears in all categories



91
92
93
94
 
# File 'lib/stuff-classifier/base.rb', line 91

def total_word_count(word)
  return 0.0 unless @word_list[word] && @word_list[word][:_total_word]
  @word_list[word][:_total_word].to_f
end

#total_word_count_in_cat(cat) ⇒ Object

return the number of words in a categories



97
98
99
100
 
# File 'lib/stuff-classifier/base.rb', line 97

def total_word_count_in_cat(cat)
  return 0.0 unless @category_list[cat] && @category_list[cat][:_total_word]
  @category_list[cat][:_total_word].to_f
end

#train(category, text) ⇒ Object

train the classifier



129
130
131
132
 
# File 'lib/stuff-classifier/base.rb', line 129

def train(category, text)
  @tokenizer.each_word(text) {|w| incr_word(w, category) }
  incr_cat(category)
end

#word_count(word, category) ⇒ Object

return number of times the word appears in a category



85
86
87
88
 
# File 'lib/stuff-classifier/base.rb', line 85

def word_count(word, category)
  return 0.0 unless @word_list[word] && @word_list[word][:categories] && @word_list[word][:categories][category]
  @word_list[word][:categories][category].to_f
end