sentiwordnet_ruby 0.0.1

data/lib/sentiwordnet_ruby.rb

@@ -0,0 +1,116 @@
+class SentiWordNet
+
+  @@sentihash = {}
+
+
+  def get_score(string)
+    sentiment_total = 0.0
+
+    #tokenize the string, also throw away some punctuation
+    tokens = tokenize(string)
+
+    tokens.each do |token|
+
+        if @@sentihash[token+"#n"]
+          sentiment_total += @@sentihash[token+"#n"]
+        end
+
+        if @@sentihash[token+"#a"]
+          sentiment_total += @@sentihash[token+"#a"]
+         end
+
+        if @@sentihash[token+"#r"]
+          sentiment_total += @@sentihash[token+"#r"]
+        end
+
+        if @@sentihash[token+"#v"]
+          sentiment_total += @@sentihash[token+"#v"]
+        end
+    end
+
+    sentiment_total
+  end
+
+  def get_sentiment(string)
+    score = get_score(string)
+
+    if (score > 0.75)
+      return "very_positive"
+    elsif (score > 0.25 && score <= 0.75)
+      return "positive"
+    elsif (score > 0 && score <= 0.25)
+      return "weak_positive"
+    elsif (score == 0)
+      return "neutral"
+    elsif (score < 0 && score >= -0.25)
+      return "weak_negative"
+    elsif (score < -0.25 && score >= -0.75)
+      return "negative"
+    elsif (score < -0.75)
+      return "very_negative"
+    end
+  end
+
+  def tokenize string
+    fold_case = false
+    string.split(/\s+/).map do |token|
+      stripped = token.gsub(/[^a-zA-Z0-9\']+/, '')
+      fold_case ? stripped.downcase : stripped
+    end
+  end
+
+
+  def self.load_defaults
+    load_senti_file(File.dirname(__FILE__) + '/SentiWordNet_3.0.0.txt')
+  end
+
+  def self.load_senti_file(filename)
+    word_temp = Hash.new
+
+    File.readlines(filename).each do |line|
+      data = line.split("\t")
+      score = data[2].to_f - data[3].to_f
+      words = data[4].split(" ")
+      words.each do |w|
+        w_n = w.split("#")
+        w_n[0] += "#"+data[0]
+        index = w_n[1].to_i - 1
+        if word_temp.has_key?(w_n[0])
+          v = word_temp[w_n[0]]
+          if (index > v.size)
+            ((v.size)..(index - 1)).each do |i|
+              v << 0.0
+            end
+          end
+          v[index] = score
+          word_temp[w_n[0]] = v
+        else
+          v = []
+          (0..(index - 1)).each do |i|
+            v << 0.0
+          end
+          v[index] = score
+          word_temp[w_n[0]] = v
+        end
+      end
+    end
+
+    word_temp.keys.each do |k|
+      word = k
+      v = word_temp[k]
+      score = 0.0
+      sum = 0.0
+      (0..(v.size - 1)).each do |i|
+        score += (1.0/(i+1).to_f)* v[i]
+      end
+      (1..(v.size)).each do |i|
+        sum += 1.to_f/i.to_f
+      end
+      score /= sum
+      @@sentihash[word] = score
+    end
+    end
+end
+
+#SentiWordNet.load_senti_file('SentiWordNet_3.0.0.txt')
+

metadata

@@ -0,0 +1,48 @@
+--- !ruby/object:Gem::Specification
+name: sentiwordnet_ruby
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+  prerelease: 
+platform: ruby
+authors:
+- Shailesh Kalamkar
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2013-04-23 00:00:00.000000000 Z
+dependencies: []
+description: A simple SentiWordNet utility for Ruby
+email: shailesh.kalamkar@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/SentiWordNet_3.0.0.txt
+- lib/sentiwordnet_ruby.rb
+- README.md
+homepage: https://github.com/kalamkar/sentiwordnet_ruby
+licenses: []
+post_install_message: 
+rdoc_options: []
+require_paths:
+- .
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 1.8.25
+signing_key: 
+specification_version: 3
+summary: SentiWordNet 3.0
+test_files: []