Upgrade to Pro — share decks privately, control downloads, hide ads and more …

ログから生まれる施策 / actions born from logs

10caa9b4d041b23c2ecd6a947fdb1607?s=47 Hiroka Zaitsu
December 13, 2016

ログから生まれる施策 / actions born from logs

WEB DIRECTORS SESSION Vol.2
http://peatix.com/event/219845

10caa9b4d041b23c2ecd6a947fdb1607?s=128

Hiroka Zaitsu

December 13, 2016
Tweet

Transcript

  1. ϖύϘͷϩάz׆༻zج൫ʮ#JHGPPUʯ ࡒ௡େՆ(.01FQBCP *OD 8&#%*3&$50344&44*0/7PM ϩά͔Βੜ·ΕΔࢪࡦ

  2. σʔλαΠΤϯςΟετσΟϨΫλʔ ࡒ௡େՆ![BJNZ NJOOFࣄۀ෦

  3. ϋϯυϝΠυϚʔέοτNJOOF IUUQTNJOOFDPN

  4. ໨࣍ w8FCαʔϏεͷߦಈϩά wϖύϘͷϩάz׆༻zج൫ʮ#JHGPPUʯ wࢪࡦ΁ͷ׆༻ྫ

  5. 8FCαʔϏεͷߦಈϩά

  6. 8FCαʔϏεͷߦಈϩά wϢʔβʔ͕αʔϏεΛར༻ͨ͠ࡍͷཤྺ wʮ͍ͭʯʮ୭͕ʯʮԿΛʯߦͬͨͷ͔ w݁Ռ͚ͩͰ͸ͳͦ͘͜ʹࢸΔաఔ΍ɺ݁Ռ్தͰఘΊͯ͠·ͬͨϢʔ βʔͷߦಈ΋෼͔Δ

  7. ߦಈϩάΛ࢖ͬͯ ΑΓྑ͍αʔϏε΁!

  8. ׆༻·Ͱͷஈ֊ wऩूߦಈϩά͕ग़ྗ͞ΕɺऔΓ·ͱΊΒΕ͍ͯΔঢ়ଶ w෼ੳऔΓ·ͱΊͨߦಈϩάΛࢹ֮Խɺ෼ੳͰ͖Δঢ়ଶ w׆༻෼ੳͨ͠ߦಈϩάΛ΋ͱʹܧଓతͳαʔϏεվળ͕ߦ͍͑ͯΔঢ়ଶ

  9. #JHGPPU IUUQTJDPOTDPN

  10. #JHGPPU wϖύϘͷϩάz׆༻zج൫ w෯޿͍δϟϯϧͷ8FCαʔϏε wϋϯυϝΠυϚʔέοτNJOOF wωοτγϣοϓ࡞੒Χϥʔϛʔγϣοϓ wϩϦϙοϓʂϨϯλϧαʔόʔ wϩάऩू͔Β׆༻·Ͱͷ֤ஈ֊ʹ͓͍ͯ൚༻తʹར༻Ͱ͖Δશࣾج൫

  11. #JHGPPU rack-bigfoot Service Request Activity log Services DB Attribute Big

    Cube Cube https://speakerdeck.com/monochromegane/pepabo-log-infrastructure-bigfoot Bandit algorithm/ Recommendation Re-targeting Feedback Name identification BI/Visualize
  12. ऩू

  13. ໊دͤ wαʔϏεͷΞΧ΢ϯτͱΫϥΠΞϯτʢ୺຤ɾϒϥ΢βʣΛඥ͚ͮ wैདྷ͸ΫϥΠΞϯτ͝ͱͷܭଌ wະϩάΠϯঢ়ଶͷΞΧ΢ϯτ΋ͦͷޙϩάΠϯͨ͠λΠϛϯάͰաڈʹ Ḫͬͯඥ͚ͮΒΕΔ wαʔϏεΛލ͍ͩඥ͚ͮ΋Մೳ

  14. ෼ੳ

  15. #JH$VCFͱ$VCF w#JHGPPU্ͷશͯͷϩάΛ#JH$VCFʹू໿ w෼ੳͷ੾ΓޱʢσΟϝϯγϣϯͱϝδϟʔʣ͕ܾ·ͬͨΒ$VCFʹूܭ wྫʣ࣌ؒ͝ͱ஫จֹۚ wߴ଎ͳࢀর͕Մೳ Activity Big Cube Cube HiveQL

    SQL Dashboard Ad-hoc query Analyst Managers, Product owners, Promotion groups
  16. ࢹ֮Խ w5BCMFBV wIUUQXXXUBCMFBVDPN wΞυϗοΫͳ෼ੳ΍ෳࡶͳσʔλ݁߹ʹ w3FEBTI wIUUQTSFEBTIJP w୭΋͕ࢀরͰ͖ΔμογϡϘʔυʹ

  17. ύε෼ੳ wίϯόʔδϣϯʹؔ͢Δ෼ੳͷ͏ͪɺܦ࿏ʢύεʣʹओ؟Λஔ͍ͨ΋ͷ wύεͷ௕࣌ؒ͞ྨܕ wΞτϦϏϡʔγϣϯϞσϧϖʔδͷग़ݱҐஔʹΑΓॏΈΛม͑Δ referrer landing last cv ??? Point

    Analytics referrer landing last Path Analytics cv
  18. ࢪࡦ΁ͷ׆༻

  19. ཭୤Ϣʔβʔ΁ͷϦςϯγϣϯ wߦಈϩά͔Β཭୤ϢʔβʔΛநग़ wΧʔτʹ౸ୡ͕ͨ͠ങΘͳ͔ͬͨϢʔβʔʢ͍ΘΏΔΧʔτམͪʣ w୹ظؒʹಉ͡࡞඼ΛԿ౓΋ݟ͍ͯΔϢʔβʔ wಛఆͷ৚݅ʹ߹க͢Ε͹ϓογϡ௨஌΍ϝʔϧͰϦςϯγϣϯ wϨγʔτϝʔϧͱಉ౳ͷ։෧཰ HiveQL Re-targeting

  20. ޿ࠂ࿈ܞ wߦಈϩά͔ΒϢʔβʔΛηάϝϯτ wϢʔβʔʹؔ࿈ੑͷߴ͍޿ࠂΛදࣔ wطʹϦʔνͨ͠Ϣʔβʔʹ͸޿ࠂΛදࣔ͠ͳ͍ʢσϦλʔήςΟϯάʣ

  21. όϯσΟοτΞϧΰϦζϜ w࡞඼ݕࡧը໘͔ΒͷརӹΛ࠷େԽ͍ͨ͠ w$53ͷҟͳΔύλʔϯͷ࡞඼ݕࡧΞϧΰϦζϜ wച্ʹର͢Δൺॏ͕ଟ͍ҝʹγεςϜมߋͷϦεΫ͕ߴ͍ w࠷ྑͷύλʔϯΛ࢖͍ͳ͕ΒΑΓྑ͍ύλʔϯΛಈతʹ୳͢

  22. όϯσΟοτΞϧΰϦζϜ w&QTJMPO(SFFEZ"MHPSJUIN w֬཰ Џ Ͱͦͷ࣌఺ͷ࠷΋ظ଴஋͕ߴ͍ύλʔϯΛ࢖༻ʢ׆༻ʣ w֬཰ЏͰϥϯμϜʹબ୒ͨ͠ύλʔϯΛ࢖༻ʢ୳ࡧʣ Activity Epsilon-Greedy algorithm User

    1-ε: exploitation ε/pattern: exploration Click / Not click Import
  23. όϯσΟοτΞϧΰϦζϜ wτϨʔυΦϑͷղܾ w׆༻͹͔Γ͍ͯ͠Δͱݱࡏͷظ଴஋ΑΓ΋ྑ͍࿹Λݟ͚ͭΒΕͳ͍ w୳ࡧ͹͔Γ͍ͯ͠Δͱظ଴஋͕ߴ͍࿹͕ར༻͞Εͳ͍

  24. Ϩίϝϯσʔγϣϯ wڠௐϑΟϧλϦϯάʹΑΔʮ͋ͳͨʹ͓͢͢Ίͷ࡞Ոʯ !NPOPDISPNFHBOFʮNJOOFNFFUT)JWFNBMMʯ IUUQTTQFBLFSEFDLDPNNPOPDISPNFHBOFQFQBCPNJOOFNBUSJYGBDUPSJ[BUJPOJOIJWFNBMM

  25. ϩάz׆༻zج൫

  26. ϩάz׆༻zج൫ wᶃਓ͕ؒ෼ੳ݁ՌΛݟͯߦ͏੩తͳʮ׆༻ʯ wࢪࡦͷਫ਼౓޲্ wᶄࣗಈԽ͞ΕͨϑΟʔυόοΫʹΑΔಈతͳʮ׆༻ʯ wͳΊΒ͔ͳγεςϜ