For the execution, we need to give two command line arguments. City Name and Input File Name. Now based on the definition of good and bad, it will create a directory by city name and put good.txt and bad.txt in that directory containing information classified as good or bad.

require 'stemmer'
require 'classifier'

#
  if ARGV.empty?

               puts "***You Should supply CityName and Input File name to script****\n"

  else  if ARGV[1]

               puts "I am searching for the city #{ARGV[0]}\n" 

               puts "The input file is #{ARGV[1]}\n"

               inputfile=ARGV[1].to_s.downcase

              pwd=Dir.getwd

              city=ARGV[0].to_s.downcase].to_s.downcase

             Dir.mkdir("#{city}")

            # Load previous classifications

            good = YAML::load_file('good.yml')

            not_good = YAML::load_file('not_good.yml')

             data=File.open("#{inputfile}","r")

            goody=File.open("#{pwd}"+"\\"+"#{city}"+"\\good.txt","a")

            nogood=File.open("#{pwd}"+"\\"+"#{city}"+"\\nogood.txt","a")

             classifier = Classifier::Bayes.new('good', 'No good')

              # Train the classifier

             not_good.each { |not_good| classifier.train_no_good not_good }

            good.each { |good_one| classifier.train_good good_one }

             while line3=data.gets

                      if classifier.classify(line3)=="Good"

                                   goody.write line3

                      else

                                   nogood.write line3

               end

        end

  else

         puts "***second argument that is name of file is required***\n"

  end

end

Quality Of Results

Quality of result depends on how much training we have given to the classifier. Its kind of a learning system where quality of result depend upon training. The major benefit of this approach is reduction in human efforts required to classify the data. Similar to this, there are many applications where human intervention is required to classify what is good and what is bad. A properly trained classifier similar to this, can be helpful in similar situations.

Hope you find this article interesting and you will be able to use it if you need to classify data for your application.

Back