Class: PheremoneAnalysis

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

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(text) ⇒ PheremoneAnalysis

Returns a new instance of PheremoneAnalysis.



18
19
20
21
22
23
24
25
26
27
28
29
 
# File 'lib/summa.rb', line 18

def initialize(document, keywords, constants)
  @pheremones = Array.new(document.wordArray.length())
  for i in 0...document.wordArray.length
    @pheremones[i] = 0
  end
  @document = document
  @keywords = keywords
  @sigma = SummaData.sigma
  @sigma_sq = @sigma * @sigma
  @threshold = SummaData.threshold
  @output = ""
end

Instance Attribute Details

#documet ⇒ Object

Returns the value of attribute documet.



163
164
165
 
# File 'lib/summa.rb', line 163

def documet
  @documet
end

#keywords ⇒ Object

Returns the value of attribute keywords.



163
164
165
 
# File 'lib/summa.rb', line 163

def keywords
  @keywords
end

#output ⇒ Object

Returns the value of attribute output.



163
164
165
 
# File 'lib/summa.rb', line 163

def output
  @output
end

#pheremones ⇒ Object

Returns the value of attribute pheremones.



163
164
165
 
# File 'lib/summa.rb', line 163

def pheremones
  @pheremones
end

#sigma ⇒ Object

Returns the value of attribute sigma.



163
164
165
 
# File 'lib/summa.rb', line 163

def sigma
  @sigma
end

#threshold ⇒ Object

Returns the value of attribute threshold.



163
164
165
 
# File 'lib/summa.rb', line 163

def threshold
  @threshold
end

Instance Method Details

#analyze ⇒ 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
 
# File 'lib/summa.rb', line 46

def analyze()
  front = 1/(@sigma * Math.sqrt(2 * Math::PI))
  for i in 0 ... @document.wordArray.length
    for j in 0 ... @keywords.wordArray.length
      if @document.docArray[i][@keywords.wordArray[j]] != nil
        for pos in 0 ... @document.wordArray.length
          temp = front * Math.exp(-((i-pos)*(i-pos))/(2*@sigma_sq))
   @pheremones[pos] = @pheremones[pos] + temp
        end
      end
    end
  end

  max_value = -1;
  for i in 0 ... @pheremones.length
    if max_value < @pheremones[i]
      max_value = @pheremones[i];
    end
  end

  for i in 0 ... @pheremones.length
    @pheremones[i] /= max_value;
  end
end

#findEndOfSentence(i) ⇒ Object 



128
129
130
131
132
133
134
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
 
# File 'lib/summa.rb', line 128

def findEndOfSentence(i)
  endIndex = @document.docArray.length
  index = i
  local_output = ""
  while(index < @document.docArray.length)
    word = @document.docArray[index]
    if(word != "Mr." && word != "Mrs." && word != "Dr." && 
	   word != "U.S." && word != "Jan." &&
	   word != "Feb." && word != "Mar." &&
	   word != "Apr." && word != "May." &&
	   word != "Jun." && word != "Jul." &&
	   word != "Aug." && word != "Sep." &&
	   word != "Oct." && word != "Nov." &&
	   word != "Dec." && word != "Sept." &&
	   word != "Lt." && word != "Maj." &&
	   word != "Col.")
      c = word[word.length - 1]
      if(c == "."[0] || c == ";"[0] || c == ":"[0])
          endIndex = index
          break;
      end
    end
    index = index + 1
  end

  if endIndex != i
    for j in i+1 ... (endIndex + 1)
      local_output += @document.docArray[j]
      local_output += " "
    end
  end

  local_output
end

#findStartOfSentence(i) ⇒ Object 



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
 
# File 'lib/summa.rb', line 96

def findStartOfSentence(i)
  index = i
  startIndex = 0
  local_output = ""
  while(index > 0)
    word = @document.docArray[index]
    if(word != "Mr." && word != "Mrs." && word != "Dr." && 
	   word != "U.S." && word != "Jan." &&
	   word != "Feb." && word != "Mar." &&
	   word != "Apr." && word != "May." &&
	   word != "Jun." && word != "Jul." &&
	   word != "Aug." && word != "Sep." &&
	   word != "Oct." && word != "Nov." &&
	   word != "Dec." && word != "Sept." &&
	   word != "Lt." && word != "Maj." &&
	   word != "Col.")
      c = word[word.length - 1]
      if(c == "."[0] || c == ";"[0] || c == ":"[0])
        startIndex = index + 1
        break;
      end
    end
    index = index - 1
  end

  for j in startIndex ... i
    local_output += @document.docArray[j]
    local_output += " "
  end
  local_output
end

#summarize ⇒ Object 



71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
 
# File 'lib/summa.rb', line 71

def summarize()
  @output = ""
  inRegion = false
  for i in 0 ... @pheremones.length
    if @pheremones[i] >= @threshold
      if inRegion
        @output += @document.docArray[i];
        @output += " ";
      else
        inRegion = true
        @output += findStartOfSentence(i)
        @output += @document.docArray[i];
        @output += " "
      end
    else
      if inRegion
        inRegion = false
        @output += findEndOfSentence(i-1)
        @output += " "
      end
    end
  end
  @output
end