toC企業でのデータ活用 (PyData.Okinawa + PythonBeginners沖縄 合同勉強会 2019)

https://pydataokinawa.connpass.com/event/80271/

Transcript

1. PyData.Okinawa + PythonBeginnersԭೄ ߹ಉษڧձ 2019 ஛໺ फ़ี (@takegue) toCاۀͰͷσʔλ׆༻; αΠΤϯεɺΤϯδχΞϦϯάͦͯ͠σβΠϯɺΞʔτ
2. 8IP
   "N
   * ஛໺ फ़ีʢ @takegue ʣ Retty ← म࢜ʢNLP; ػց຋༁ʣˡ ߴઐ
 Core Value: Data Architect 
 σʔλͷՁ஋Λ࠷େԽ͢Δ࢓૊Έ/ઃܭͷ࣮ݱ
 
 ࣥච׆ಈ: 
 ʮ༏ઌ౓ֶशʹΑΔਪનจ͔Βͷݟग़͠நग़ʯ
 ʮ΍ͬͯΈΑ͏ʂ ػցֶशʢSotware Designʣʯ
 ʮࢼֶͯ͠Ϳ ػցֶशೖ໳ʯଞ…
 
 ߴ౓AIਓࡐ͔΋ʁ
 ͦͷଞ: https://shwca.se/takegue
3. Data Architectͷ͓͠͝ͱ: ྲྀ௨ͱσʔλͷܦࡁݍΛ࡞Δ͜ͱ ΞφϦετ σʔλϚʔτ ϓϩμΫτ σʔλ΢ΣΞϋ΢ε
4. σʔλʹؔΘΔδϣϒɺ͍ΖΜͳδϣϒ͕͋Δ • Data Scientists • Data Infrastructure Engineer • ML Engineer / SysML • BI Engineer / Data Platform Engineer • Data Visualization Engineer / Data Analyst • Data Application Engineer
5. ܅͸ͲΜͳδϣϒʹͳΓ͍ͨʁʂ
6. ͜ͷઌੜ͖࢒ΔͨΊʹ͜͏ͳΓ͍ͨ ʮ͜ͷձࣾͷ໋ߝ͸Զ͕Ѳ͍ͬͯΔʯʢը૾ུʣ ݴͬͯΈͨ͘ͳ͍ʁ
7. toCاۀͱͯ͠ͲͷΑ͏ʹσʔλ׆༻ʹऔΓ૊ΜͰ͍Δͷ͔ʁ
8. ·ͣ͸αʔϏε঺հ: Retty
10. https://retty.me/announce/philosophy/
12. σʔλʹؔΘΔλεΫɺ͍ΖΜͳλεΫ͕͋Δ
13. σʔλʹؔΘΔλεΫɺ͍ΖΜͳλεΫ͕͋Δ ΞϓϦ πʔϧ
14. σʔλʹؔΘΔλεΫɺ͍ΖΜͳλεΫ͕͋Δ ΞϓϦ πʔϧ ૊৫ن໛Ͱͷ
 εέʔϧϝϦοτ͕ߴ͍ Ϣʔβن໛Ͱͷ
 εέʔϧϝϦοτ͕ߴ͍
15. σʔλʹؔΘΔλεΫɺ͍ΖΜͳλεΫ͕͋Δ ந৅త ۩ମత ΞϓϦ πʔϧ
16. σʔλʹؔΘΔλεΫɺ͍ΖΜͳλεΫ͕͋Δ ந৅త ۩ମత ΞϓϦ πʔϧ ௕ظˍܧଓత։ൃ޲͚ʢR&Dʣ ୹ظత/ूதత։ൃ ςί͕ޮ͖΍͍͢; 
 3ഒͷੜ࢈ੑˠ 10ഒͷ੒ՌʹมΘͬͨΓ͢Δ 10ഒͷੜ࢈ੑͷҧ͍͕ͦͷ··10ഒͷ੒Ռͷࠩ
17. σʔλʹؔΘΔλεΫɺ͍ΖΜͳλεΫ͕͋Δ ந৅త ۩ମత ΞϓϦ πʔϧ Ϩίϝϯυ ݕࡧ ίϯςϯπੜ੒ ଟݴޠରԠ ޿ࠂ ίϯςϯπ؂ࢹ ؂ࢹ (ҟৗݕ஌౳) ऩӹ༧ଌ ࣗಈQA ୳ࡧతσʔλ෼ੳ ϝτϦΫε։ൃ ύϑΥʔϚϯε෼ੳ Ծઆݕূ
18. σʔλʹؔΘΔλεΫɺ͍ΖΜͳλεΫ͕͋Δ ந৅త ۩ମత ΞϓϦ πʔϧ Ϩίϝϯυ ίϯςϯπੜ੒ ଟݴޠରԠ ޿ࠂ ίϯςϯπ؂ࢹ ؂ࢹ (ҟৗݕ஌౳) ऩӹ༧ଌ ࣗಈQA ୳ࡧతσʔλ෼ੳ ϝτϦΫε։ൃ ύϑΥʔϚϯε෼ੳ Ծઆݕূ ݕࡧ
19. ػցֶशϓϩΤΫτͷҰྫ: ʮ༏ઌ౓ֶशʹΑΔਪનจ͔Βͷݟग़͠நग़ʯ
 from http://www.orsj.or.jp/archive2/or62-11/or62_11_731.pdf
20. ΩϟονίϐʔΛࣗಈతʹܾΊΕΔΑ͏ʹ͍ͨ͠ ୺తʹ͍͏ͱ
 ͓ళͷͨΊͷΩϟονίϐʔΛ࡞Δ
21. ͸͡Δͱ͖ʹԿΛߟ͑Δ͔ʁ •ΤϯδχΞϦϯάγϯΩϯά (ٕज़తʹͳΜͱ͔͢Δ) •ϓϩμΫτγϯΩϯά (Ϣʔεέε/Ձ஋ਫ४ΛఆΊΔ) •αΠΤϯεγϯΩϯά (໰୊ͷຊ࣭Λ໰͏)
22. ϓϩμΫτγϯΩϯά • ඼࣭ج४ͱͯ͠ʮ৴པʯΛଛͳΘͳ͍͔ʁ • ӕΛ͔ͭͳ͍͜ͱ • ޱޠతͰ͸ͳ͍ͳͲʮΒ͠͞ʯΛද͢બ޷ੑ͕͋Δ • αʔϏεͷڧΈʹͳΔΑ͏ͳ΋ͷ͕๬·͍͠ • Ωϟονίϐʔͱͯ͠ͷཱͪҐஔ; ັྗతͳจͰ͋Δ͜ͱ • ͋ͨΓ͞ΘΓͷͳ͍ฏۉతͳจষΛٻΊ͍ͯΔΘ͚Ͱ͸ͳ͍
 • ϓϩμΫτ΁ͷ౷߹ͷ਌࿨ੑ͕ߴ͍͜ͱ͕๬·͍͠ • จࣈ਺੍ݶͷ໰୊ (PC΍εϚϗ)
23. αΠΤϯεγϯΩϯά • ັྗతʹײ͡Δจͱ͸Կ͔ʁ ◦ධՁͷઃܭ ▪ આಘྗ͕͋ΔΩϟονίϐʔ → CTR্͕Γͦ͏ʁ ▪ CTR͕͕͋ΔΩϟονίϐʔ ≠ આಘྗʁ
 • ΩϟονίϐʔΒ͍͠ͱ͸Կ͔ʁ ◦Ωϟονίϐʔͷྲྀெੑ ≠ จͱͯ͠ͷྲྀெੑ ◦จͱͯ͠͸ଟগ่Ε͍ͯͯ΋ྑ͍ʢϦζϜ͕͋Δͱྑ͍ʣ ◦ʮ͜ͷ͓ళͷεύήοςΟ͸ඒຯ͍͠Ͱ͢ʯ ◦ʮඒຯͳεύήοςΟΛఏڙʂʯ • Ωϟονίϐʔ͸ʮޱίϛʯͷཁ໿ͳͷ͔ • ͦ΋ͦ΋NLPͱͯ͠͸Ͳ͜·Ͱ͕Ͱ͖Δൣғͳͷ͔ʁ ◦ ػցతʹྲྀெͳจΛੜ੒͢Δ͜ͱ͸Ͱ͖Δ͔ʁ ▪ ػցֶशόοΫάϥ΢ϯυͱͯ͠ͷ஌ݟ ◦ Ͳ͏͍͏࣮ݧઃఆͩͬͨΒ͏·͘ਐΊΒΕΔ͔ʁ
24. ΤϯδχΞϦϯάʹ͜ΕΒΛ౿·͑ͯͳΜͱ͔͢Δ
     • ϓϩμΫτΠϯͷखؒ͸ʁ ◦ DBʹಥͬࠐΜͰͪΐͬͱίʔυΛॻ͖׵͑Δ͚ͩɺ͓खܰ؆୯ʂ ◦ ࢼߦࡨޡͷํʹ͕͖࣌ؒ͞΍͍͢ • ࠷ߴਫ਼౓ͷख๏͕ඞཁͳ༁Ͱ͸ͳ͍ ◦ ख๏ࣗମʹ৽نੑ͕ͳͯ͘ྑ͍ɻ஌ݟͷ৽نੑ͸ཉ͍͠ ◦ ࢼͯ͠ධՁͯ͠վળͰ͖Δ΋ͷ͕ྑ͍ ◦ ֶशʹ͕͔͔࣌ؒΔ௒େن໛ֶश͸࠷ॳ͸΍Βͳ͍ • ݱঢ়͋Δσʔληοτͷ೺Ѳ ◦ Ωϟονίϐʔͷจ਺͸͔ͳΓ͋Δ (20ສจڧ) ˍ ޱίϛ΋ͨ͘͞Μ͋Δʂ ◦ ੜ੒͢ΔͨΊʹ׬શʹ੔උ͞Εͨσʔληοτ͸ͳ͍ ˍ ୹ظܾઓ (1.0ϱ݄) ◦
25. ྫ͑͹ … • ςϯϓϨʔτࢤ޲ ◦ ௒େྔͷൈ͚͕݀͋ΔςϯϓϨʔτΛ༻ҙ͠
 ٖࣅతʹେྔͷจΛੜ੒͠ɺͦ͜ͷத͔Βྑ͍΋ͷΛબͿ ▪ ΩϟονίϐʔͷݴޠϞσϧͰྲྀெੑ͸ධՁͰ͖Δʂ ▪ ΩϟονίϐʔͷςϯϓϨʔτΛ͍͔ʹఏڙͰ͖Δ͔ʁ • ׬શจੜ੒ࢤ޲ ◦ GANGAN͍͜͏ͥʂ ◦ ΍ͬͨ͜ͱͳ͍͠ɺ΍ָͬͯͯͦ͠͏
 • ޱίϛཁ໿ࢤ޲ ◦ ޱίϛΛཁ໿੍ͯ͠ݶ͞ΕͨจࣈͰจΛͭ͘Δ
26. Ͳ͏͔ͨ͠ʁ • ཁ໿ʢநग़ʣࢤ޲ͷΞϓϩʔνͱͯ͠໰୊ΛϞσϧԽ • ̎ͭͷจʹରͯ͠ɺࣄྫؒͷॱংؔ܎>= Λֶश͢Δ2஋෼ྨثΛߏங͢Δ໰୊ͱͯ͠ϞσϧԽ ɹɹɹ 
 ྑ͍ΩϟονίϐʔΛઈରతͳࢦඪͰܭଌ͢Δͷ͸೉͍͕͠ ɹɹɹ ૬ରతͳؔ܎͸؆୯ʹఆٛͰ͖Δɻ ɹɹɹ ɹɹɹॱংؔ܎͕ఆٛͰ͖Δͱιʔτ͕Ͱ͖Δʂ f(X1 , X2 ) = F(ϕ(X1 ) − ϕ(X2 ))) = { 1, if X1 ≥ X2 0, otherwise f(“͜ͷ͓ళͷຯḩो͸͏·͍”, “ࣗՈ੡ͷຯḩो͸͓;͘Ζͷຯʂ”ʣ
 = “͜ͷ͓ళͷຯḩो͸͏·͍” =< “ࣗՈ੡ͷຯḩो͸͓;͘Ζͷຯʂ"
27. Ͳ͏͔ͨ͠ʁ • Ωϟονίϐʔͷจ >= ޱίϛ͔ΒϥϯμϜʹ੾Γग़ͨ͠จɹͰେྔͷٖࣅσʔλΛੜ੒ ɹɹ େྔͷ܇࿅ࣄྫˍग़ྗͷ࣍ݩ਺2Ͱ͋ΔͨΊɺֶशͰ͖ͦ͏ͳؾ͕͢Δ ΦϯϥΠϯߋ৽͕ՄೳͳϩδεςΟοΫճؼΛ෼ྨثʹར༻͢Δ͜ͱͰ
 ɹɹ σʔλྔʹରͯ͠΋໰୊ͳֶ͘शͰ͖ΔΑ͏ʹ (sklearn.linear_model.SGDClassifier Λར༻) ɹɹ Ұ؏ੑͷ͋Δσʔλྔ͕े෼ʹ֬อͰ͖Δͱ NNܥͷػցֶश͸ɺ͍͍ͩͨͲΜͳࣸ૾Ͱ΋Ͱ͖Δ ৄࡉ͸ׂѪ (http://www.orsj.or.jp/archive2/or62-11/or62_11_731.pdf)
28. Ͳ͏͔ͨ͠ʁ ◦ ྑ͍ޱίϛ͔Βྑ͍Ωϟονίϐʔ͕͓ళʹ০ΒΕΔʂ • Ϣʔβͷޱίϛ͕͓ళΛԠԉ͢Δͱ͍͏ɺαʔϏεͷՁ஋؍ͱ΋Ϛον ◦ ॊೈੑ͕ߴ͍: ਪ࿦ϑΣʔζͷࡍͷจͷੜ੒ํ๏Λ޻෉͢Ε͹ɺ
 ৭ʑͳύλʔϯͰΩϟονίϐʔ͕ੜ੒Ͱ͖Δ ◦ ղऍੑ΋ߴ͍: Ϟσϧ͕ͱͯ΋୯७ͳͨΊ ▪ ϩδεςΟοΫճؼͷಛ௃ྔͷॏΈΛ෼ੳ͢Ε͹ ▪ ୯ޠ-unigram: Ωϟονίϐʔʹ࢖ΘΕ΍͍͢ಛ௃తͳ୯ޠ͕Θ͔Δ ▪ ୯ޠ-ngram: จମֶ͕शͰ͖Δɻະ஌ޠॲཧΛߦ͏͜ͱͰςϯϓϨʔτ΋֫ಘͰ͖Δɻ ◦ ੜ੒͢ΔͷͰ͸ͳ͘ ධՁثΛ࡞͍ͬͯΔͷͰɺΦϖϨʔγϣϯʹରͯ͠਌࿨ੑ͕ߴ͍ ▪ Ϋϥ΢υιʔγϯάͰ͋Ε͹ɺॳֶऀͷ܇࿅ʹ࢖͑Δ ▪ ੒Ռ෺ͷϑΟϧλͱͯ͠ͷԠ༻΋ߟ͑ΒΕΔ
29. ݁Ռ: Ͳ͏͍͏Ωϟονίϐʔ͕Ͱ͖Δ͔ʁ ࣾ಺ͰͷਓखධՁͰ͸ ఆྔతʹ΋ਓ͕ؒ࡞੒ͨ͠ΑΓ༗ҙʹྑ͍Ωϟονίϐʔ͕Ͱ͖Δ͜ͱ͕Θ͔ͬͨ શళฮͰ͸ແཧ͕ͩಛఆͷϑΟϧλΛ͔·ͤ͹ϓϩμΫτΠϯ΋Ͱ͖ͨ ◦ (ਓख) ౎಺࠷ڧͷ͏ͲΜ ◦ (ػց) ே͔Β൩·Ͱ௕ऄͷྻ͕Ͱ͖Δ໊ళ͏ͲΜ԰͞Μ ◦ (ػց) ೋށ࢈ͷͦ͹Λళ಺Ͱ੡ค͠ɺṢ͖ͨͯɾଧͪͨͯɾᣐͰͨͯͷʮ̏ͨͯʯͰఏڙ ◦ (ਓख) ͓ംͪΌΜͷՈʹ༡ͼʹདྷͨΑ͏ͳݹຽՈͰ௖͘ίγͷڧ͍͓ڶഴ͸ඒຯ
30. ݁Ռ: Ͳ͏͍͏Ωϟονίϐʔ͕Ͱ͖Δ͔ʁ ࣾ಺ͰͷਓखධՁͰ͸ ఆྔతʹ΋ਓ͕ؒ࡞੒ͨ͠ΑΓ༗ҙʹྑ͍Ωϟονίϐʔ͕Ͱ͖Δ͜ͱ͕Θ͔ͬͨ શళฮͰ͸ແཧ͕ͩಛఆͷϑΟϧλΛ͔·ͤ͹ϓϩμΫτΠϯ΋Ͱ͖ͨ ◦ (ਓख) ౎಺࠷ڧͷ͏ͲΜ ◦ (ػց) ே͔Β൩·Ͱ௕ऄͷྻ͕Ͱ͖Δ໊ళ͏ͲΜ԰͞Μ ◦ (ػց) ೋށ࢈ͷͦ͹Λళ಺Ͱ੡ค͠ɺṢ͖ͨͯɾଧͪͨͯɾᣐͰͨͯͷʮ̏ͨͯʯͰఏڙ ◦ (ਓख) ͓ംͪΌΜͷՈʹ༡ͼʹདྷͨΑ͏ͳݹຽՈͰ௖͘ίγͷڧ͍͓ڶഴ͸ඒຯ
31. toCྖҬͰͷσʔλ׆༻ʢػցֶशʣͷ஌ݟ
     ྑ͍σʔλ͸໰୊Λγϯϓϧʹͯ͘͠ΕΔ ▪ ྑ͍໰୊ઃఆ͸ෳ਺ͷղܾΛ༩͑ͯ͘ΕΔ (Simple > Easy) ▪ ʢαʔϏεʗۀքʗλεΫʣυϝΠϯಛ༗ͷಛԽ͢Δ͜ͱͰɺΑΓ໰୊ΛγϯϓϧʹͰ͖Δ Ұఆਫ४ͷ୲อʹͱͯ΋ۤ࿑͢Δ ◦ ϞσϧʙγεςϜͷ͏·͍ύΠϓϥΠϯͱͯ͠ͷઃܭྗ͕ࢼ͞ΕΔ ◦ ΞΧσϛοΫͰ͋Ε͹ ͻͱͭͣͭͰධՁɾղܾ͢Δෳ਺ͷ໰୊Λಉ࣌ʹղܾ͢Δඞཁ͕͋Δ ◦ Ωϟονίϐʔͷ৔߹͸ ྲྀெੑ / ৴པੑʢղऍੑʣ / ॊೈੑ Λಉ࣌ʹຬͨ͢ඞཁ͕͋ͬͨ ◦ ਓͷؒҧ͍ʹ͸ൺֱతڐ༰త͕ͩɺػցతͳؒҧ͍͸ඇڐ༰త ʢਓΈ͍ͨʹؒҧ͍͑ͨʣ A/BςετͷΑ͏ͳܗͰΠϯϋ΢εͳධՁ͕ར༻Ͱ͖ΔΞυόϯςʔδ ◦ αʔϏεಛ༗ͷ݁Ռʹͳͬͯ͠·͏ͨΊɺଞͷαʔϏεʹ͓͍ͯͷ࠶ݱੑ͸୲อͰ͖ͳ͍͕…
32. toCྖҬͰͷσʔλ׆༻ʢػցֶशʣͷ೰Έ
     ໘ന͍ྖҬͰ΋͋Δ
33. toCͱͯ͠ޮՌతʹσʔλΛར༻͢ΔͨΊʹ͸Ͳ͏͢΂͖͔ʁ
34. ͍͔ʹσʔλ׆༻Λߦ͏͔ʁ from https://simplystatistics.org/2019/04/17/tukey-design-thinking-and-better-questions/ ղ͚Δ໰୊ͷ
 ೉қ౓ ղ͘΂͖໰୊ͷ
 ඼࣭
35. ͍͔ʹσʔλ׆༻Λߦ͏͔ʁ NNͷ୆಄ from https://simplystatistics.org/2019/04/17/tukey-design-thinking-and-better-questions/ ղ͚Δ໰୊ͷ
 ೉қ౓ ղ͘΂͖໰୊ͷ
 ඼࣭
36. ͍͔ʹσʔλ׆༻Λߦ͏͔ʁ NNͷ୆಄ ਅͷGOAL from https://simplystatistics.org/2019/04/17/tukey-design-thinking-and-better-questions/ ղ͚Δ໰୊ͷ
 ೉қ౓ ղ͘΂͖໰୊ͷ
 ඼࣭
37. ͍͔ʹσʔλ׆༻Λߦ͏͔ʁ NNͷ୆಄ ਅͷGOAL ཧ૝ from https://simplystatistics.org/2019/04/17/tukey-design-thinking-and-better-questions/ ղ͚Δ໰୊ͷ
 ೉қ౓ ղ͘΂͖໰୊ͷ
 ඼࣭
38. ͍͔ʹσʔλ׆༻Λߦ͏͔ʁ NNͷ୆಄ ਅͷGOAL ཧ૝ ͜͜ʹ͍Δͭ΋Γʁ from https://simplystatistics.org/2019/04/17/tukey-design-thinking-and-better-questions/ ղ͚Δ໰୊ͷ
 ೉қ౓ ղ͘΂͖໰୊ͷ
 ඼࣭
39. ͍͔ʹσʔλ׆༻Λߦ͏͔ʁ NNͷ୆಄ ਅͷGOAL ཧ૝ ͜͜ʹ͍Δͭ΋Γʁ ࣮ࡍ͸͔ͬͪ͜΋ʁ from https://simplystatistics.org/2019/04/17/tukey-design-thinking-and-better-questions/ ղ͚Δ໰୊ͷ
 ೉қ౓ ղ͘΂͖໰୊ͷ
 ඼࣭
40. ͍͔ʹσʔλ׆༻Λߦ͏͔ʁ ղ͚Δ໰୊ͷ
 ೉қ౓ ղ͘΂͖໰୊ͷ
 ඼࣭ NNͷ୆಄ ਅͷGOAL ཧ૝ ͜͜ʹ͍Δͭ΋Γʁ ࣮ࡍ͸͔ͬͪ͜΋ʁ from https://simplystatistics.org/2019/04/17/tukey-design-thinking-and-better-questions/ ΪϟοϓΛຒΊΔ ྑ͍໰͍Λߟ͑Δඞཁ͕͋Δ
41. ྑ͍σʔλ͕͋Ε͹໰୊͸γϯϓϧʹͳΔ … ͱ͢Δͱ ྑ͍σʔλΛ͍͔ʹ࡞Δ͔Λߟ͑ΔͨΊʹ
 ςΫϊϩδʔ΍ΤϯδχΞϦϯά͚ͩͰͳ͘ ྑ͍σʔλ͕ಘΒΕΔαΠΫϧΛߟ͑Δͱྑͦ͞͏
42. Krebs Cycle of Creativity https://jods.mitpress.mit.edu/pub/AgeOfEntanglement
43. ཧ૝ͷʮਓ޻஌ೳʯ͸ϧϯό…ʁ https://twitter.com/atochotto/status/1129183985119051776? ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed&ref_url=https%3A%2F%2Fpaperusercontent.com%2Fintegrations%2Fembed%2Fiframe%2Ftweet%3Fid%3D1129183985119051776 จԽΛม͑ΒΕΔʮਓ޻஌ೳʯ͸੒ޭ ϧϯόͷத਎͕ػցֶशͷ࢓༷ͷ༗ແ͸
 ۃ࿦Ͳ͏Ͱ΋͍͍͔΋͠Εͳ͍
44. Case Study: Google຋༁ ਓ޻஌ೳܥͷαʔϏε͕ʮσβΠϯʯ͔ΒʮจԽʯʹӨڹ༩͑ΔྫΛߟ࡯ͯ͠Έ͍ͨͱࢥ͏ • ̏೥͙Β͍લ͔ΒΊͪΌΑ͘ͳͬͨɻ • ͪΐ͏Ͳػց຋༁ͷύϥμΠϜ͕େ͖͘มΘΔλΠϛϯάΛ • ΞΧσϛοΫଆͰݟ͍ͯͨͷͰɺػց຋༁Ͱͷ୊ࡐΛ͋͛ͯΈ͍ͨ
45. Case Study: Google຋༁ (ௌऺͷօ͞Μʹ࣭໰) • ຋༁ීஈ࢖͍͍ͯ͠ΔਓɺखΛ͋͛ͯΈͯཉ͍͠ ◦ தֶߍ΍ߴߍͰͷ॓୊ͷࡍʹར༻͍ͯͨ͠Γ ◦ Θ͔Βͳ͍୯ޠΛࣙॻ͕ΘΓʹࡧҾ͢Δਓ ◦ Πϯλʔωοτ αʔϑΟϯͰӳޠͷχϡʔεΛ຋༁͢Δ༻్ ◦ શ͘࢖ͬͯͳ͍ਓ
46. Case Study: Google຋༁ ωλཁһʁ
47. Case Study: Google຋༁ from NLP2017νϡʔτϦΞϧʮθϩ͔Β࢝ΊΔ χϡʔϥϧωοτϫʔΫػց຋༁ʯ (http://lotus.kuee.kyoto-u.ac.jp/~nakazawa/NLP2017-NMT-Tutorial.pdf) ৽͍͠ൃ໌ or ٕज़తʹઌߦ͍ͯͨ͠ͷ͔ʁ
48. Krebs Cycle of Creativity https://jods.mitpress.mit.edu/pub/AgeOfEntanglement
49. Google຋༁͸ϓϩμΫτͱͯ͠ͳͥ੒ޭ͍ͯ͠Δ͔ʁ Α͏ͳؾ͕͢Δ
50. ڊਓͨͪ͸ͲΜͳࢹ఺Λ͍࣋ͬͯΔͷ͔
51. ڊਓͨͪͷಈ͖ʹண໨ͯ͠ΈΔ
52. ڊਓͨͪͷಈ͖ʹண໨ͯ͠ΈΔ - Netflix
53. ڊਓͨͪͷಈ͖ʹண໨ͯ͠ΈΔ - Netflix: byDevTools
    
54. ڊਓͨͪͷಈ͖ʹண໨ͯ͠ΈΔ - Netflix: ίϯςϯπ৘ใ
55. ڊਓͨͪͷಈ͖ʹண໨ͯ͠ΈΔ - Netflix: Πϕϯτ৘ใ - ϩέʔϧ৘ใ - ࢪࡦ൪߸৘ใ - ϦϦʔε൪߸৘ใ - Ͳͷίϯςϯπ͕ݟΒΕͨ - Ͳͷίϯςϯπ͕දࣔ͞Ε͔ͨ - ͲͷλΠϛϯάͰ
 ίϯςϯπ͕ಈ͍͔ͨ …
56. ڊਓͨͪͷಈ͖ʹண໨ͯ͠ΈΔ - AirBnB
57. ڊਓͨͪͷಈ͖ʹண໨ͯ͠ΈΔ - AirBnB https://www.slideshare.net/mounialalmas/tutorial-on-online-user-engagement-metrics-and-optimization • Ϣʔβߦಈʹؔ͢ΔମܥɺͲͷΑ͏ʹଌఆ͍͔͕ͯ͘͠·ͱ·͍ͬͯΔ
58. ڊਓͨͪͷಈ͖ʹண໨ͯ͠ΈΔ - AirBnB https://www.kdd.org/kdd2018/accepted-papers/view/real-time-personalization-using-embeddings-for-search-ranking-at-airbnb ϢʔβͷৼΔ෣͍͔Βݕࡧ݁ՌʢίϯςϯπʣΛ࠷దԽ͢Δ: ηογϣϯϩάΛ Skip-gramϞσϧͰղऍͯ͠ɺίϯςϯπͱͯ͠ͷྨࣅ౓Λߏங → ݕࡧ݁Ռʹ൓ө
59. ݸਓͱͯ͠ΑΓޮՌతͳσʔλ׆༻Λ͍ͯͨ͘͠Ίʹ
60. υϝΠϯΛཧղ͢Δ͜ͱɺҙຯͷ͋ΔAIΛ࡞Δ͜ͱ͸࿈ଓత
61. Data Architectͷ͓͠͝ͱ: ྲྀ௨ͱσʔλͷܦࡁݍΛ࡞Δ͜ͱ ΞφϦετ σʔλϚʔτ ϓϩμΫτ σʔλ΢ΣΞϋ΢ε
62. RettyͰͷ׆ಈ͸͜ΕΒΛͭͳ͛Δ׆ಈ
63. ϢʔβΛཧղ͢Δ͜ͱɺ໾ʹཱͭAIΛ࡞Δ͜ͱ͸࿈ଓత σʔλ΁ͷؔΘΓํ͸΋ͬͱ৭ʑ͍͍͋ͬͯ
64. ·ͱΊ σʔλ͸Ұੜ๞͖ͳ͍
65. ͓ΘΓ