Class: ActivityMapper::Linguistics::Tagger

Inherits:
Object
  • Object
show all
Defined in:
lib/activity_mapper/linguistics.rb

Constant Summary collapse

UNINTERESTING_PARTS_OF_SPEECH = 
['DT', 'PRP', 'IN', 'CC', 'MD']

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ Tagger

Returns a new instance of Tagger.



17
18
19
20
21
22
23
24
25
 
# File 'lib/activity_mapper/linguistics.rb', line 17

def initialize()
  @lexicon = {}
  file = File.new(File.join(File.dirname(__FILE__), 'lexicon.txt'), 'r')
  file.each_line {|line|
                    toks=line.split
                    @lexicon[toks.shift]=toks
                 }
  file.close
end

Class Method Details

.keywords_for_caption(caption) ⇒ Object 



31
32
33
34
35
36
37
38
39
40
41
42
43
44
 
# File 'lib/activity_mapper/linguistics.rb', line 31

def self.keywords_for_caption(caption)
  @@tagger ||= self.new
  keywords = []
  all_keywords = @@tagger.tokenize(caption) 
  pos_tags = @@tagger.part_of_speech_tag(all_keywords)
  all_keywords.each_with_index do |keyword,i|
    next if UNINTERESTING_PARTS_OF_SPEECH.include?(pos_tags[i])
    next unless keyword.size > 4
    keywords << keyword.downcase
  end
  keywords
rescue => e
  all_keywords
end

Instance Method Details

#part_of_speech_tag(text) ⇒ Object 



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
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
 
# File 'lib/activity_mapper/linguistics.rb', line 46

def part_of_speech_tag(text)

  ## start by tokenizing strings passed in
  if text.class == String then
    text = tokenize(text) 
  end

  ## we only work on arrays.  If text isn't an array, 
  ## quit now.
  if text.class != Array then
    raise RuntimeError, "can't tokenize #{text.class}"
  end

  # this looks like an artifact of testing
  #	puts "text:",text,"\n"
  ret = []
    
  text.each do
    |w| ret << (@lexicon[w] && @lexicon[w][0]) || 
    (@lexicon[w.downcase] && words[w.downcase][0]) || 
      'NN' 
  end
    
  ## Now, apply transformational rules:
  text.length.times do |i|

    ## rule 1: DT, {VBD | VBP} --> DT, NN
    if i > 0 then
      if ret[i - 1] == "DT" then
        if ret[i] == "VBD" or ret[i] == "VBP" or ret[i] == "VB" then
          ret[i] = "NN"
        end
      end
    end

    ## rule 2: convert a noun to a number (CD) if "." appears in the word
    if ret[i] =~ /^N/ then
      if text[i] =~ /\./ then
        ret[i] = "CD"
      end
    end

    ## rule 3: convert a noun to a past participle if words[i] ends 
    ## with "ed"
    if ret[i] =~ /^N/ && text[i] =~ /ed$/ then
      ret[i] = "VBN"
    end

    ## rule 4: convert any type to adverb if it ends in "ly"
    if text[i] =~ /ly$/ then
      ret[i] = "RB"
    end

    ## rule 5: convert a common noun (NN or NNS) to a adjective if 
    ## it ends with "al"
    if ret[i] =~ /^NN/ && text[i] =~ /al$/ then
      ret[i] = "JJ"
    end

    ## rule 6: convert a noun to a verb if the preceeding work is "would"
    if i > 0 then
      if ret[i] =~ /^NN/ then
        if text[i-1].downcase == "would" then
          ret[i] = "VB"
        end
      end
    end

    ## rule 7: if a word has been categorized as a common noun and 
    ## it ends with "s", then set its type to plural common noun (NNS)
    if ret[i] == "NN" && text[i] =~ /s$/ then
      ret[i] = "NNS"
    end

    ## rule 8: convert a common noun to a present participle 
    ## verb (i.e., a gerand)
    if ret[i] =~ /^NN/ && text[i] =~ /ing$/ then
      ret[i] = "VBG"
    end

    ## rule 9: <noun> <noun 2> --> <noun> <verb> if <noun 2> 
    ## can also be a verb
    if i > 0  then
      if ( ( ret[i-1] =~ /^NN/ ) && ( ret[i] =~ /^NN/ ) )  then
        if @lexicon[text[i]].include?("VBN")  then
          ret[i] = "VBN"
        end
        if @lexicon[text[i]].include?("VBZ") then
          ret[i] = "VBZ"
        end
      end
    end

 end
 return ret
end

#tokenize(words) ⇒ Object 



27
28
29
 
# File 'lib/activity_mapper/linguistics.rb', line 27

def tokenize(words)
  words.split(/ |,|\.|\:|\;|\'/) #'
end