Unsupervised topic modeling in Ruby using LDA

gistfile1.rb

require 'lda-ruby'
corpus = Lda::Corpus.new
corpus.add_document(Lda::TextDocument.new(corpus, "a lion is a wild feline animal", []))
corpus.add_document(Lda::TextDocument.new(corpus, "a dog is a friendly animal", []))
corpus.add_document(Lda::TextDocument.new(corpus, "a cat is a feline animal", []))

lda = Lda::Lda.new(corpus)
lda.verbose = false
lda.num_topics = (2)
lda.em('random')
topics = lda.top_words(3)

# Results
# => {0=>["animal", "friendly", "dog"], 1=>["animal", "feline", "cat"]} 

LDA is short for Latent Dirichlet Allocation, an algorithm developed by David Blei, at Princeton. It is an unsupervised mechanism for surfacing topics and themes within a group of text.