Last active
March 3, 2019 17:02
-
-
Save Innarticles/99f884bd0f6a03ad112df49ad0c733ba to your computer and use it in GitHub Desktop.
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
max = 0.0 | |
if !QaCorp.where("question like ?", "%#{self.text}%").empty? | |
# Exact question exist in the database. | |
answer = QaCorp.where("question like ?", "%#{self.text}%").first.answer | |
max = 1.0 | |
else | |
corpus = QaCorp.all.map {|qa| TfIdfSimilarity::Document.new(qa.question)} | |
d_question = TfIdfSimilarity::Document.new(self.text) | |
corpus << d_question | |
model = TfIdfSimilarity::TfIdfModel.new(corpus, :library => :nmatrix) | |
matrix = model.similarity_matrix | |
puts "calculating matrix" | |
fixed_text = "" | |
puts "searching the nearest question-anser pair" | |
corpus.each do |doc| | |
# score is a number between 0.0 and 1.0 | |
if d_question.text != doc.text | |
score = matrix[model.document_index(d_question), model.document_index(doc)] | |
puts "score is #{score}" | |
if score > max | |
max = score | |
fixed_text = doc.text | |
end | |
end | |
end | |
if max > 0.7 #make this a hyper parameter for the model | |
answer = QaCorp.where("question like ?", "%#{fixed_text}%").first.answer | |
else | |
answer = nil | |
end | |
end | |
puts max | |
answer |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment