W2vUtils.jl

Talk about my first Julia package presented in JuliaTokyo #3 on 2015-04-25.

Transcript

1. W2vUtils.jl Kenta Murata 2015-04-25 JuliaTokyo #3
2. W2vUtils.jl Kenta Murata 2015-04-25 JuliaTokyo #3
3. introduce(self)
4. Kenta Murata @mrkn ଜా ݡଠ ✓ Cookpad Inc. ✓ Ruby committer as
 a bigdecimal maintainer ✓ One of Julia beginners
5. Julia and me
6. Julia and me ✓ First meet is 2012? in Wikipedia
7. Julia and me ✓ First meet is 2012? in Wikipedia ✓ First use is 2013 only in REPL
8. Julia and me ✓ First meet is 2012? in Wikipedia ✓ First use is 2013 only in REPL ✓ First script writing is yesterday!!
9. W2vUtils.jl
10. W2vUtils.jl
11. W2vUtils.jl ✓ Utilities for word2vec data
12. W2vUtils.jl ✓ Utilities for word2vec data ✓ My first Julia package
13. W2vUtils.jl ✓ Utilities for word2vec data ✓ My first Julia package ✓ My first Julia script
14. Why W2vUtils.jl?
15. Why W2vUtils.jl? ✓ I’ve tried to write SOM learner at first for this presentation
16. Why W2vUtils.jl? ✓ I’ve tried to write SOM learner at first for this presentation ✓ What the most interesting application for SOM?
17. Why W2vUtils.jl? ✓ I’ve tried to write SOM learner at first for this presentation ✓ What the most interesting application for SOM? ✓ I think it is interesting to map distributed representations of words onto 2d-lattice.
18. But…
19. But… ✓ I couldn’t get to done to write SOM learner
20. But… ✓ I couldn’t get to done to write SOM learner ✓ Writing both data loader and SOM learner is too many to done in one night
21. But… ✓ I couldn’t get to done to write SOM learner ✓ Writing both data loader and SOM learner is too many to done in one night ✓ So I completely focused on to make my first package
22. http://www.slideshare.net/KentaSato/julia-36649709
23. https://github.com/mrkn/W2vUtils.jl
24. using W2vUtils
25. Install W2vUtils.jl Pkg.clone("git@github.com:mrkn/W2vUtils.jl.git")
26. Load word2vec data using W2vUtils wv = load("vectors.bin", W2vData) nwords(wv) #=> The number of words vocabulary(wv) #=> The array of words projdim(wv) #=> Dimensions of vector (== 200) projection(wv) #=> The projection matrix wordindex(wv, word) #=> Lookup index of the word wordindices(wv, words) #=> Lookup indices of the words
27. http://techlife.cookpad.com/entry/2015/02/27/093000
28. N-best nearest words
29. N-best nearest words
30. N-best nearest words using W2vUtils wv = load("recipe_steps.bin", W2vData) (words, dists) = distance(wv, "νϣί"; n=5) collect(zip(words, dists)) #=> 5-element Array{(UTF8String,Float64),1}: ("νϣίϨʔτ",0.9378173408020709) ("Ψφογϡ",0.7568368811212932) ("ϚγϡϚϩ",0.7461657278585042) ("Ϗλʔνϣί",0.7439865689272069) ("Ϋϥϯν",0.7296649975102198)
31. Word analogy
32. Word analogy
33. Word analogy using W2vUtils wv = load("recipe_steps-phrase.bin", W2vData) (words, dists) = analogy(wv, ["໊ݹ԰", "੺ຯḩ", "௕໺"]; n=5) collect(zip(words, dists)) #=> 5-element Array{(UTF8String,Float64),1}: ("৴भ",0.5091896142078718) ("ຯḩ",0.5033710820183608) ("ഴ_ຯḩ",0.5033459461705277) ("৴भ_ຯḩ",0.49885121244487496) ("੺ຯḩ_നຯḩ",0.4816681271563857)
34. Nearest words for a vector using W2vUtils wv = load("recipe_steps-phrase.bin", W2vData) ੺ຯḩ = projection(wv, "੺ຯḩ") ໊ݹ԰ = projection(wv, "໊ݹ԰") ௕໺ = projection(wv, "௕໺") (words, dists) = nearest_words(wv, ੺ຯḩ - ໊ݹ԰ + ௕໺; n=5) collect(zip(words, dists)) #=> 5-element Array{(UTF8String,Float64),1}: ("௕໺",0.6527766969046891) ("੺ຯḩ",0.5241494934036854) ("৴भ",0.5091896142078718) ("ຯḩ",0.5033710820183608) ("ഴ_ຯḩ",0.5033459461705277)
35. Example usage
37. ͜Ε΍Γ͍ͨ
38. examples/w2v_pca.jl using W2vUtils using MultivariateStats using Gadfly wv = load(ARGS[1], W2vData) (words, dists) = distance(wv, ARGS[2]; n=15) vecs = W2vUtils.projection(wv, words) model = fit(PCA, vecs'; maxoutdim=2) transvecs = transform(model, vecs') pca_plot = plot(x=transvecs[1, :], y=transvecs[2, :], label=words, Geom.point, Geom.label) draw(PDF(ARGS[3], 4inch, 3inch), pca_plot)
39. � ���� ��� ���
      
     
         
       
    
     ���� ���� ��� ��� ��� ��� � $ julia w2v_pca.jl recipe_steps.bin νϣί pca1.pdf
41. ͜Ε΋΍ͬͯΈΑ͏
42. $ julia w2v_pca.jl recipe_steps-phrase.bin ੺ຯḩ pca2.pdf � ���� ��� ��� ���
    �  �
    ��  � �  � �  � ��  � �
    � ���� ��� ��� �
43. Future work
44. Future work
45. Future work ✓ Conform to the standard coding style of Julia
46. Future work ✓ Conform to the standard coding style of Julia ✓ Submit to METADATA.jl
47. Future work ✓ Conform to the standard coding style of Julia ✓ Submit to METADATA.jl ✓ Implement self-organizing map
48. Future work ✓ Conform to the standard coding style of Julia ✓ Submit to METADATA.jl ✓ Implement self-organizing map ✓ Visualize 2d-lattice map of word distributed representations