tonality_analyser 0.0.1

data/.gitignore

@@ -0,0 +1,17 @@
+*.gem
+*.rbc
+.bundle
+.config
+.yardoc
+Gemfile.lock
+InstalledFiles
+_yardoc
+coverage
+doc/
+lib/bundler/man
+pkg
+rdoc
+spec/reports
+test/tmp
+test/version_tmp
+tmp

data/Gemfile

@@ -0,0 +1,4 @@
+source 'https://rubygems.org'
+
+# Specify your gem's dependencies in tonality_analyser.gemspec
+gemspec

data/LICENSE.txt

@@ -0,0 +1 @@
+Copyright (c) 2013 Samuel Sanchez

data/README.md

@@ -0,0 +1,29 @@
+# TonalityAnalyser
+
+  Process text and propose tonality.
+
+## Installation
+
+Add this line to your application's Gemfile:
+
+    gem 'tonality_analyser'
+
+And then execute:
+
+    $ bundle
+
+Or install it yourself as:
+
+    $ gem install tonality_analyser
+
+## Usage
+
+You can read spec files
+
+## Contributing
+
+1. Fork it
+2. Create your feature branch (`git checkout -b my-new-feature`)
+3. Commit your changes (`git commit -am 'Add some feature'`)
+4. Push to the branch (`git push origin my-new-feature`)
+5. Create new Pull Request

data/Rakefile

@@ -0,0 +1,8 @@
+require "bundler/gem_tasks"
+
+require 'rspec/core/rake_task'
+
+task :default => :spec
+desc 'Run tests with RSpec'
+RSpec::Core::RakeTask.new(:spec)
+

data/lib/tonality_analyser.rb

@@ -0,0 +1,3 @@
+require 'tonality_analyser/version'
+require 'tonality_analyser/helpers'
+require 'tonality_analyser/engine'

data/lib/tonality_analyser/engine.rb

@@ -0,0 +1,77 @@
+module TonalityAnalyser
+
+  # Refactor: work to Redis !
+  #   Redis ! Redis ! Redis ! Redis ! Redis ! :)
+  class Engine
+    TONALITIES = [:pos, :neg]
+    attr_reader :counted_words, :probabilites
+    def initialize
+      @total_words = {}
+      @total_words[:all] = 0
+      @total_words[:pos] = 0
+      @total_words[:neg] = 0
+      @counted_words = {}
+      @counted_words[:pos] = {}
+      @counted_words[:neg] = {}
+      @probabilites = {}
+      @probabilites[:pos] = {}
+      @probabilites[:neg] = {}
+      @spec_probabilites = {}
+      @spec_probabilites[:pos] = {}
+      @spec_probabilites[:neg] = {}
+    end
+    def train(words, tonality)
+      raise "Invalid tonality '#{tonality}'" unless TONALITIES.include?(tonality)
+      words.split.each do |w|
+        word = Helpers::Text.normalize(w)
+        @total_words[:all] += 1
+        @counted_words[tonality][word] = @counted_words[tonality].include?(word) ? @counted_words[tonality][word]+1 : 1
+      end
+    end
+    def compute_probabilities!
+      # TODO: Refactor this :)
+      @counted_words[:pos].each do |word, count|
+        @probabilites[:pos][word] = @counted_words[:pos][word].to_f / (@counted_words[:pos][word].to_f + @counted_words[:neg][word].to_f)
+        @spec_probabilites[:pos][word] = @probabilites[:pos][word]
+      end
+      @counted_words[:neg].each do |word, count|
+        @probabilites[:neg][word] = @counted_words[:neg][word].to_f / (@counted_words[:pos][word].to_f + @counted_words[:neg][word].to_f)
+        @spec_probabilites[:neg][word] = @probabilites[:neg][word]
+      end
+    end
+    def analysis(text, tonality)
+      num, den1, den2 = 1.0, 1.0, 1.0
+
+      words = Helpers::Text.clean_words_from(text)
+      words.each do |word|
+        @spec_probabilites[tonality][word] ||= 0.01
+        num *= @spec_probabilites[tonality][word]
+      end
+      num *= 0.5
+      words.each do |word|
+        @probabilites[tonality][word] ||= 0.01
+        den1 *= @probabilites[tonality][word]
+      end
+      words.each do |word|
+        den2 *= (1 - @probabilites[tonality][word])
+      end
+      proba_pol = num / (den1 + den2)
+      proba_pol = 0.0 if proba_pol.nan?
+      proba_pol
+    end
+    def tonality(text)
+      pos_proba = analysis(text, :pos)
+      neg_proba = analysis(text, :neg)
+      pos_proba >= neg_proba ? :pos : :neg
+    end
+    def load_traning_corpus!
+      TONALITIES.each { |tonality| load_traning_corpus_for(tonality) }
+    end
+    def load_traning_corpus_for(tonality)
+      File.open("./training/#{tonality}.txt", 'r') do |f|
+        f.each_line { |line| train(line, tonality) }
+        f.close
+      end
+    end
+  end
+end

data/lib/tonality_analyser/helpers.rb

@@ -0,0 +1,17 @@
+module TonalityAnalyser
+  module Helpers
+
+    class Text
+      def self.normalize(word)
+        word.downcase.gsub(/[^0-9a-z]/i, '')
+      end
+      def self.clean_words_from(text)
+        text.downcase.gsub(/[^0-9a-z]/i, ' ').split.inject([]) do |words, w|
+          words << w if w.length > 2
+          words
+        end
+      end
+    end
+
+  end
+end

data/lib/tonality_analyser/version.rb

@@ -0,0 +1,3 @@
+module TonalityAnalyser
+  VERSION = "0.0.1"
+end

data/spec/spec_helper.rb

@@ -0,0 +1,3 @@
+require 'rspec'
+$:.unshift(File.dirname(__FILE__) + '/../lib')
+require 'tonality_analyser'

data/spec/tonality_analyser_spec.rb

@@ -0,0 +1,15 @@
+require 'spec_helper'
+
+describe TonalityAnalyser::Engine do
+  
+  # Woouuu Wouuuu les test de la mort :)
+  it 'propose tonality' do
+    e = TonalityAnalyser::Engine.new
+    e.load_traning_corpus!
+    e.compute_probabilities!
+    e.tonality('This').should == :neg
+    e.tonality('Unacceptable') == :neg
+    e.tonality('want') == :pos
+    e.tonality('Thanks') == :pos
+  end
+end

data/tonality_analyser.gemspec

@@ -0,0 +1,19 @@
+# -*- encoding: utf-8 -*-
+lib = File.expand_path('../lib', __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+require 'tonality_analyser/version'
+
+Gem::Specification.new do |gem|
+  gem.name          = "tonality_analyser"
+  gem.version       = TonalityAnalyser::VERSION
+  gem.authors       = ["Samuel Sanchez"]
+  gem.email         = ["samuel@pagedegeek.com"]
+  gem.description   = %q{Process text and propose tonality.}
+  gem.summary       = %q{Process text and propose tonality with bayes computation}
+  gem.homepage      = "http://github.com/PagedeGeek/tonality_analyser"
+
+  gem.files         = `git ls-files`.split($/)
+  gem.executables   = gem.files.grep(%r{^bin/}).map{ |f| File.basename(f) }
+  gem.test_files    = gem.files.grep(%r{^(test|spec|features)/})
+  gem.require_paths = ["lib"]
+end

data/training/neg.txt

@@ -0,0 +1,58 @@
+I hate this.
+This sucks.  I can't even figure out how to open it.
+Wow, this was horrible. Nothing is worse.
+Very bad.  Very ugly.  Very slow.
+Whoever wrote this should die in a fire.
+This is the crappiest crap that ever crapped.
+This was ugly, despicable, and horrid.  Imprecise yet slow.
+I loathed it.  I wanted to set it on fire.
+The company should be ashamed of releasing this.  Shame!
+I hated it when I saw it, I hate it now, I will hate it tomorrow.
+Loathsome and disgusting.  Why is it slimy?
+This is just a scam.  Inaccurate and incorrect.
+Ick - repulsive and unworthy of my money.
+Expensive yet declasse.  Just screams nouveau riche.
+I abhorred it.  It's horrible.  Just horrible.
+I found it appalling.  Why would anyone buy such a thing?
+My entire family hated it.
+The entire family found it beastly and cruel to animals.
+My entire family hated its disgusting noises.
+Noisy.  They don't tell you how loud it is.
+Detestable.  It makes a disagreeable whir, clank, grind noise.
+I bought one for my sister, and it was detestable.
+Eerie noises came from its dreadful orifices.
+Throughout the night, it made obnoxious noises.
+Atrocious quality - never buying again.
+This company should be ashamed of the bad news coming out.
+Inferior quality and awful craftsmanship.
+Cruddy and cheap pieces.  Looks pretty blah, too.
+Looks like a careless monkey put it together.
+Where was QA?  This crummy thing didn't even last two days.
+Defective in oh so many ways.  Godawful piece of crap.
+I was unsatisfied with my purchase.
+Just seeing it is a downer.
+Had seven faulty sensors and nine faulty gauges.
+Inadequate lighting and slimy, junky interior.
+Gross!  It spit at me!  I want to spit at it!
+Slipshod, shoddy craftsmanship.
+Dishonest purveyors of snake oil.
+Just pure garbage.  It belongs in a dumpster or a landfill.
+Poor quality and cheapness is evident.
+Substandard for something of this level.  I paid too much.
+I paid way too much for this piece of crap.
+It's the pits.  Unacceptable and unsatisfactory.
+This had a lousy exterior and an inferior interior.
+The worst product I've ever used in my life.  Never again!
+I would compare it unfavorably to feces.
+It is teh suck.  It sucks more than a sucky suck.
+Worse than anything I can comprehend.
+Hellish, boorish, and vain.
+Shoddy worksmanship.  Fell apart in the rain.
+Whoever made this will have to answer to his maker.
+Broke down almost immediately.  Broken since then.
+Hate hate hate hate hate hate hate hate hate hate hate hate hate hate!
+Crap crap crap crap crap crap crap crap crap crap crap crap crap crap!
+Sucky suck suck sucked sucking sucked really bad.
+Bad bad bad bad bad bad bad bad bad bad bad bad bad bad bad bad bad!
+Dislike dislike dislike dislike dislike dislike dislike dislike dislike!
+Lousy lousy lousy lousy lousy lousy lousy lousy lousy lousy lousy lousy!

data/training/pos.txt

@@ -0,0 +1,59 @@
+This is awesome!
+I love this.  It has the right mixture of color and texture.
+Love love love.  I can't go to sleep without it.
+My child loves it.  She says it's her favorite toy.
+I can't even begin to describe how great this is.  Thank you!
+Thanks to the developers who made this.  Pure awesome sauce.
+I just want to snuggle in this program and fall asleep.
+It's cuter than a bug in a rug.
+Technical excellence, achieved.
+The company releases pure, 100%, grade-A greatness, every year.
+I've said it before, and I'll say it again - this is the best.
+The absolute best product I've ever used.
+Grade A.. no, A+!  Would buy again.
+First-class seating at a reasonable price.
+Quite admirable.  Adroitly handles every exception.
+Marvelous.  As soft as a fluffy bunny but still tough.
+The last item was transcendent.  I could see time.
+A consummate perfectionist, he has created a better mousetrap.
+Absolutely perfect in every conceivable way.
+It surpassed all of my expectations, including some I didn't know about.
+100% brilliance.  More brilliant than a diamond.
+Not just good - downright exceptional.
+Fine taste put together with expert hands.
+You can tell an expert swordsmith made this.
+Entirely complete - an utter enchantment to work with.
+I found it good - utterly good - a champion among wines.
+Super-duper, out of sight, groovy!  An out and out perfect game.
+Not just fantastic - super-duper fantastic.
+Hellacious - a heavy mixture of nitroglycerin and excitement.
+The terrific end of summer experience was wonderful.
+Wonderful, wonderful, wonderful.  So wonderful.
+A superb, virtuoso performance.  Distinguished for excellence.
+Outstanding in its field.  Makes me want to jump for joy.
+Oh yes, I like it, I love it, I want some more of it.
+This is a great way to store your cherries in the winter.
+It's the cat's pajamas, the way it accurately knows things.
+Quite hunky-dory - even peachy keen.
+Peachy-keen.  It's a choice celebration.
+Exquisite craftsmanship and meritorious conduct.
+Quite simply, it's the greatest thing ever.
+A-1.  A1.  A-OK.  Just A+!
+Crackerjack skills under it all, but a smooth finish on top.  Neat!
+Wow, this was cool.  The coolest thing ever.  Cooler than ice.
+Beautiful.  Attractive.  It fit her like a glove.
+When I dream about it, it's wonderful.  Works like a dream.
+Commendable in their attention to detail.
+I found it deserving of the highest awards in its field, and more.
+Worthy of the Nobel prize, the Turing Award, and probably an Emmy.
+No ifs, ands, or buts - solid build with an unreal body.
+This is just awesome.  Awesome to the max!  Rad!
+I found everything all right (alright).  Superior to doing it by hand.
+Definitely praiseworthy - state of the art.
+Wow, that was slick.  Wow!
+Very nice, very pleasant.  As smooth as butter.
+Beautiful.  Awe-inspiring.  Awesome.
+Like like like like like like like like like like like like!
+Love love love love love love love love love love love love !
+Awesome awesome awesome awesome awesome awesome awesome awesome !
+Cool cool cool cool cool cool cool cool cool cool cool cool cool !

metadata

@@ -0,0 +1,61 @@
+--- !ruby/object:Gem::Specification
+name: tonality_analyser
+version: !ruby/object:Gem::Version
+  version: 0.0.1
+  prerelease: 
+platform: ruby
+authors:
+- Samuel Sanchez
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2013-02-10 00:00:00.000000000 Z
+dependencies: []
+description: Process text and propose tonality.
+email:
+- samuel@pagedegeek.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- .gitignore
+- Gemfile
+- LICENSE.txt
+- README.md
+- Rakefile
+- lib/tonality_analyser.rb
+- lib/tonality_analyser/engine.rb
+- lib/tonality_analyser/helpers.rb
+- lib/tonality_analyser/version.rb
+- spec/spec_helper.rb
+- spec/tonality_analyser_spec.rb
+- tonality_analyser.gemspec
+- training/neg.txt
+- training/pos.txt
+homepage: http://github.com/PagedeGeek/tonality_analyser
+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: Process text and propose tonality with bayes computation
+test_files:
+- spec/spec_helper.rb
+- spec/tonality_analyser_spec.rb