Last active
August 29, 2015 13:57
-
-
Save sasamijp/9913369 to your computer and use it in GitHub Desktop.
ミキ的にはこのコード、エロいものかそうでないかのスコアを出すには全然合ってないって思うな
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# -*- encoding: utf-8 -*- | |
require 'natto' | |
@natto = Natto::MeCab.new | |
def getSS | |
sentenses = [] | |
File.read("./ss.txt", :encoding => Encoding::UTF_8).split("\n").each do |str| | |
sentenses.push str if str.include?("「") and str.include?("」") | |
end | |
sentenses.map! do |sentense| | |
sentense = sentense.split("「")[1].chop | |
end | |
return sentenses | |
end | |
def getSSnoEro | |
sentenses = [] | |
File.read("./ss_notero.txt", :encoding => Encoding::UTF_8).split("\n").each do |str| | |
sentenses.push str if str.include?("「") and str.include?("」") | |
end | |
sentenses.map! do |sentense| | |
sentense = sentense.split("「")[1].chop | |
end | |
return sentenses | |
end | |
def extractnouns(str) | |
nouns = [] | |
@natto.parse(str) do |word| | |
case word.feature.split(",")[0] | |
when '名詞' | |
nouns.push(word.surface) | |
end | |
end | |
return nouns | |
end | |
allnouns = [] | |
allnouns_ero = [] | |
allnouns_notero = [] | |
getSS.push(getSSnoEro).flatten.compact.each do |sentense| | |
allnouns.push extractnouns sentense | |
end | |
getSS.compact.each do |sentense| | |
allnouns_ero.push extractnouns sentense | |
end | |
getSSnoEro.compact.each do |sentense| | |
allnouns_notero.push extractnouns sentense | |
end | |
def getWordscore(word, allnouns, allnouns_ero, allnouns_notero) | |
found = 0 | |
allnouns_ero.each do |noun| | |
if noun.include? word | |
found += 1 | |
end | |
end | |
occuranceInNegatives = 0 | |
allnouns_notero.each do |noun| | |
if noun.include?(word) | |
occuranceInNegatives += 1 | |
end | |
end | |
wordAppearInEro = found.to_f / allnouns_ero.length | |
prior = allnouns_ero.length.to_f / allnouns.length.to_f | |
occuranceInAll = found + occuranceInNegatives | |
wordAppearInDocuments = occuranceInAll.to_f / allnouns.length.to_f | |
if wordAppearInDocuments == 0 | |
return nil | |
end | |
if occuranceInAll >= 5 | |
confidence = 1.0 | |
else | |
confidence = occuranceInAll * 0.2 | |
end | |
return [prior * wordAppearInEro / wordAppearInDocuments, confidence] | |
end | |
def dumpAllScore(allnouns, allnouns_ero, allnouns_notero) | |
allnouns.each do |noun| | |
score = getWordscore(noun.to_s, allnouns, allnouns_ero, allnouns_notero) | |
if score != nil | |
puts "#{noun},#{score[0]},#{score[1]}\n" | |
end | |
end | |
end | |
dumpAllScore(allnouns.flatten, allnouns_ero, allnouns_notero) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment