Get the text similarity you need with word embeddings

A 5 minute talk at PyData London on 7 Feb.

Transcript

1. Get the word similarity you need
   Lev Konstantinovskiy
   Community Manager at Gensim
   @teagermylk
   http://rare-technologies.com/
   

   View Slide

2. Streaming
   We turn NLP papers into industrial Python code.
   

   View Slide

3. Credits
   Parul Sethi
   Undergraduate student
   University of Delhi, India
   RaReTech Incubator program
   Added WordRank to Gensim
   http://rare-technologies.com/incubator/
   

   View Slide

4. Business Problems
   

   View Slide

5. Business Problems
   “What does Elizabeth think about Mr Darcy?”
   “Male characters in Pride and Prejudice?”
   

   View Slide

6. Two Different
   Business Problems
   1) What words are in the topic of “Darcy”?
   2) What are the Named Entities in the text?
   

   View Slide

7. P&P is only 120k words
   

   View Slide

8. Closest word to “king”?
   Trained on Wikipedia 17m words
   Attribute Interchangeable Both
   

   View Slide

9. Tensorflow has awesome viz!
   

   View Slide

10. How to get the similarity you need
    My similar words must
    be
    Associated Interchangeable
    I want to describes the
    word’s
    Topic Function
    I want to Know what doc is about Recognize names
    Then I should run Wordrank (even on small
    corpus, 1m words)
    or
    Word2vec skipgram big
    window needs large corpus
    >5m words
    Word2vec skipgram small
    window
    or
    FastText
    or
    VarEmbed
    

    View Slide

11. Rare and Frequent words are
    incomprehensible
    

    View Slide

12. Thanks!
    Lev Konstantinovskiy
    github.com/tmylk
    @teagermylk
    Gensim T-shirt question:
    How many words are in
    Pride and Prejudice?
    

    View Slide