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

ECサイトにおける閲覧履歴を用いた購買に繋がる行動の変化検出 / Change Detecti...

ECサイトにおける閲覧履歴を用いた購買に繋がる行動の変化検出 / Change Detection in Behavior Followed by Possible Purchase Using Electronic Commerce Site Browsing History

財津大夏, 三宅悠介
GMOペパボ株式会社 ペパボ研究所
2020.05.15 第49回 情報処理学会 インターネットと運用技術研究会

Hiroka Zaitsu

May 15, 2020
Tweet

More Decks by Hiroka Zaitsu

Other Decks in Technology

Transcript

  1. ࡒ௡େՆ, ࡾ୐༔հ / Pepabo R&D Institute, GMO Pepabo, Inc. 2020.05.15

    ୈ49ճ ৘ใॲཧֶձ Πϯλʔωοτͱӡ༻ٕज़ݚڀձ ECαΠτʹ͓͚ΔӾཡཤྺΛ༻͍ͨ ߪങʹܨ͕ΔߦಈͷมԽݕग़
  2. • ϢʔβʔͷߦಈͷมԽʹ߹ΘͤͯECαΠτͷγεςϜΛదԠతʹ มԽͤ͞Δ͜ͱͰߪങ཰ͷ޲্͕ظ଴͞ΕΔ • ঎඼Λ୳ࡧ͍ͯ͠Δ ➡ ଟ༷ੑͷ͋Δਪનख๏ʹ੾Γସ͑ͯڵຯΛऒ͘ • ಛఆ঎඼ͷߪങΛߦ͓͏ͱ͍ͯ͠Δ ➡

    ܾࡁಋઢΛࣔͯ͠ߪങΛଅ͢ • ECαΠτͷγεςϜͷదԠతͳมԽΛ࣮ݱ͢ΔͨΊʹɼ Ϣʔβʔ͕ԿΒ͔ͷߦಈΛऔͬͨ௚ޙʹมԽΛݕग़͍ͨ͠ 5 Ϣʔβʔͷߦಈʹ߹ΘͤͨECαΠτͷదԠతͳมԽ
  3. • طଘݚڀʹ͓͚ΔʮϢʔβʔͷ໨తʹରԠ͢ΔӾཡύλʔϯͷ෼ྨʯ(*1,2) • ॳظஈ֊ɿΧςΰϦʔϖʔδͱ঎඼ϖʔδΛଟ͘Ӿཡ͢Δ • ߪങͷ௚લɿগ਺ͷ঎඼ϖʔδʹӾཡ͕ूத͢Δ • Ϣʔβʔ͝ͱͷ͋Δظؒͷʮ঎඼Ӿཡճ਺ʯͱʮ঎඼ͷछྨͷ਺ʯ͸ ࣍ݩ਺ͷগͳ͍ಛ௃ྔʹͳΓ͏Δ *1

    Moe, W.W.: Buying, searching, or browsing: Differentiating between online shoppers using in-store navigational clickstream, Journal of Consumer Psychology, Vol.13, Is-sues 1-2, pp.113-123 (2003). *2 ΢Οϥϫϯɾυχɾμϋφ:৘ใ୳ࡧͷ໨తΛߟྀͨ͠ߪങܾఆϞσϧ,ϚʔέςΟϯάɾαΠΤϯε, Vol.25, No.1,pp.15-35 (2017). 9 طଘݚڀ͔Βͷಛ௃ྔͷީิ
  4. • ঎඼ͷӾཡճ਺ʹର͢Δ঎඼ͷଐੑͷछྨͷൺ • Ϣʔβʔ ͷߦಈཤྺ • ʹ͸঎඼Ӿཡ ΍঎඼ݕࡧ ͳͲ͕͋Δ •

    ͷ೚ҙͷҐஔͷ΢Οϯυ΢ Λߟ͑Δ • ୠ͠ɼ΢Οϯυ΢αΠζ ͱ ͔ͭ Λຬͨ͢࠷খͷࣗવ਺ Λ༻͍ͯ u Su = (a1 , a2 , …, al ) a aview asearch Su Wu (t) = (a′ 1 , a′ 2 , a′ 3 , …, at ) w 1 < n < w t − w + n > 0 n a′ 1 = at−w+n a′ 2 = at−w+n+1 a′ 3 = at−w+n+2 14 ಛ௃ྔͷఆٛᶃ
  5. • ঎඼ͷӾཡճ਺ʹର͢Δ঎඼ͷଐੑͷछྨͷൺ • ͷ೚ҙͷҐஔͷ΢Οϯυ΢ ʹ͓͚Δ • ঎඼ͷଐੑ ͷछྨʹؔ͢Δू߹ Λ༻͍ͯ ಛ௃ྔ

    • ஋͕খ͍͞΄Ͳߪങʹ޲͔͍ͬͯΔ Su Wu (t) = (a′ 1 , a′ 2 , a′ 3 , …, at ) aview ͷର৅ͱͳͬͨ঎඼ͷଐੑ attr ͷछྨ ঎඼ͷӾཡ aview ͷճ਺ attr rattr(Wu (t)) = || count(aview) 15 ಛ௃ྔͷఆٛᶄ
  6. • Ϣʔβʔɹͷߦಈཤྺ • Ͱͷ঎඼IDʹؔ͢Δಛ௃ྔ • ͱ ͷର৅ͷ঎඼ID=1ɼ ͷର৅ͷ঎඼ID=2ͱ͢Δͱ Su =

    (asearch 1 , aview 2 , aview 3 , asearch 4 , aview 5 , aview 6 , aview 7 , aview 8 , aview 9 , apurchase 10 ) Wu (5) = (asearch 1 , aview 2 , aview 3 , asearch 4 , aview 5 ) aview 2 aview 3 aview 5 rID(Wu (5)) = || count(aview) = 2 3 16 ಛ௃ྔͷྫ u Wu (5)
  7. • ಛ௃ྔͷਪҠͷ΢Οϯυ΢ Λߟ͑Δ • ୠ͠ɼ΢Οϯυ΢αΠζ ͱ ͔ͭ Λຬͨ͢࠷খͷࣗવ਺ Λ༻͍ͯ(*) •

    Λ೚ҙͷ఺Ͱೋ෼ͨ͠΢Οϯυ΢ ͱ ʹରͯ͠ ౷ܭతԾઆݕఆʹΑΔฏۉ஋ͷࠩͷݕఆΛద༻ • ༗ҙਫ४ Ͱ༗ҙࠩ͋Γͱݟͳͨ͠৔߹ʹ ͷ࠷ॳͷཁૉΛมԽ఺ͱݟͳ͢ * r' ΛٻΊΔࣜΛݚڀใࠂͷ࣌఺͔Βमਖ਼͍ͯ͠·͢ W′ u (t) = (r′ 1 , r′ 2 , r′ 3 , …, rattr(Wu (t))) w′ 1 < m < w′ t − w′ + m > 0 m r′ 1 = rattr(Wu (t − w′ + m)) r′ 2 = rattr(Wu (t − w′ + m + 1)) r′ 3 = rattr(Wu (t − w′ + m + 2)) W′ u (t) W′ 1 W′ 2 s W′ 2 17 ಛ௃ྔͷਪҠΛ༻͍ͨมԽݕग़ͷఆٛᶃ
  8. • ౷ܭతԾઆݕఆʹΑΔฏۉ஋ͷࠩͷݕఆʹ͸ Welch ͷ ݕఆΛ༻͍Δ • Student ͷ ݕఆͷվྑ •

    ฼෼ࢄ͕౳͍͜͠ͱΛԾఆ͠ͳ͍ • ෼෍ͷ࿪ΈʹରԠ͕Մೳ • ඪຊͷ฼෼ࢄ͕౳͘͠ͳ͍৔߹ʹ΋޿ൣʹରԠ͠͏Δ t t 18 ಛ௃ྔͷਪҠΛ༻͍ͨมԽݕग़ͷఆٛᶄ
  9. • ͷͱ͖ ͷ֤૊ʹ Welch ͷ ݕఆΛద༻ • ͱ ͷ૊Ͱ༗ҙࠩ͋Γͱݟͳͨ͠৔߹ ͷ࣌ࠁ

    ΛมԽ఺ͱݟͳ͢ W′ u (t) = (r′ 1 , r′ 2 , r′ 3 , r′ 4 , r′ 5 ) W′ 1 = (r′ 1 ) W′ 2 = (r′ 2 , r′ 3 , r′ 4 , r′ 5 ) W′ 1 = (r′ 1 , r′ 2 ) W′ 2 = (r′ 3 , r′ 4 , r′ 5 ) W′ 1 = (r′ 1 , r′ 2 , r′ 3 ) W′ 2 = (r′ 4 , r′ 5 ) W′ 1 = (r′ 1 , r′ 2 , r′ 3 , r′ 4 ) W′ 2 = (r′ 5 ) t W′ 1 = (r′ 1 , r′ 2 ) W′ 2 = (r′ 3 , r′ 4 , r′ 5 ) r′ 3 = rattr(Wu (t − w′ + m + 2)) t 19 ಛ௃ྔͷਪҠΛ༻͍ͨมԽݕग़ͷྫ
  10. • ECαΠτʮminneʯͷϓϩμΫγϣϯ؀ڥʹ͓͚ΔӾཡཤྺ • 2020೥3݄1೔0͔࣌Β24࣌·Ͱͷσʔλ • Ӿཡཤྺ ͷܥྻ௕ ͷ 96,984 Ϣʔβʔ

    • ൺֱͷͨΊߪങΛߦͬͨϢʔβʔͱߦΘͳ͔ͬͨϢʔβʔʹ෼ׂ • ࡞඼ʹඥͮ͘4ͭͷଐੑͰ࣮ݧ • ࡞඼IDɼ࡞඼ͷग़඼ऀIDɼ࡞඼ͷΧςΰϦάϧʔϓɼ࡞඼ͷΧςΰϦ Su l ≥ 6 σʔληοτ 22
  11. ϋΠύʔύϥϝʔλͷݕ౼ • Ӿཡཤྺ͔Βಛ௃ྔͷ஋ΛٻΊΔࡍͷ΢Οϯυ΢ͷ෯ Λ {5,10} Ͱ࣮ݧ • ಛ௃ྔͷ஋ͷมԽΛݕग़͢Δࡍͷ΢Οϯυ΢ͷ෯ Λ {3,5}

    Ͱ࣮ݧ • ߪങϢʔβʔʹؔͯ͠ΑΓଟ͘ͷมԽ఺Λݕग़͠ɼඇߪങϢʔβʔʹؔͯ͠ গͳ͍มԽ఺Λݕग़ͨ͠ ͱ ΛҎ߱ͷ࣮ݧʹ༻͍ͨ • ༗ҙਫ४ • ׳ྫతͳ஋ͱͯ͠ Λ༻͍ͨ w w′ w = 10 w′ = 5 s s = 0.05 24
  12. • ࡞඼ଐੑ͝ͱͷಛ௃ྔͷ஋ͷਪҠΛശͻ͛ਤͰ֬ೝ • ྫ ఏҊख๏ʹద͢Δ࡞඼ଐੑͷߟ࡯ 25 • ԣ࣠ɿ࣌ܥྻ • ॎ࣠ɿಛ௃ྔͷ஋

    • ശͷ্୺ɿୈࡾ࢛෼Ґ਺ • ശͷԼ୺ɿୈҰ࢛෼Ґ਺ • ശͷதͷԣઢɿதԝ஋ • ͻ͛ͷ্୺ɿୈࡾ࢛෼Ґ਺ʴ࢛෼Ґൣғͷ1.5ഒ • ͻ͛ͷԼ୺ɿୈҰ࢛෼Ґ਺−࢛෼Ґൣғͷ1.5ഒ • ͻ͛ͷ্Լͷ఺ɿ֎Ε஋ • ੺͍ॎઢɿதԝ஋ʹରͯ͠ఏҊख๏Λద༻ͯ͠ݕग़ͨ͠มԽ఺
  13. ӅΕϚϧίϑϞσϧʢHMMʣͱͷൺֱᶄ • ఏҊख๏ΑΓ΋HMMͷํ͕ੵۃతʹʮߪങϢʔβʔʯͷϥϕϧΛ෇͚ͨ ࡞඼IDΛಛ௃ྔʹ༻͍ͨ৔߹ͷࠞಉߦྻ ਖ਼ղϥϕϧ ߪങ ඇߪങ ༧ଌϥϕϧ ఏҊख๏ ߪങ

    526 4551 ඇߪങ 201 4245 HMM ߪങ 662 5571 ඇߪങ 65 3225 ࡞඼ͷग़඼ऀIDΛಛ௃ྔʹ༻͍ͨ৔߹ͷࠞಉߦྻ ਖ਼ղϥϕϧ ߪങ ඇߪങ ༧ଌϥϕϧ ఏҊख๏ ߪങ 483 5719 ඇߪങ 244 3077 HMM ߪങ 679 7047 ඇߪങ 48 1749 29
  14. ӅΕϚϧίϑϞσϧʢHMMʣͱͷൺֱᶅ • ఏҊख๏ • ਅͷඇߪങϢʔβʔʹର͢Δਫ਼౓͕ߴ͍ • ِཅੑ཰ʹରِͯ͠ӄੑ཰͕௿͍ • ߪങʹܨ͕ΔϢʔβʔͷߦಈͷมԽݕग़ͷ໨తʹԊ͍ͬͯΔ •

    HMM • ਅͷߪങϢʔβʔʹର͢Δਫ਼౓͕ߴ͍ • ʮߪങ͠ͳ͔ͬͨʯʹϚοϐϯά͞ΕΔӅΕঢ়ଶͷ෼෍͕ฏۉ1.0ɼඪ४ภࠩ1.16*10−8ͱͳͬͯ ͓Γɼ͔ᷮͰ΋ಛ௃ྔͷ஋͕ݮগ͢Δͱʮߪങͨ͠ʯӅΕঢ়ଶʹભҠ͍ͯͨ͠ 30
  15. ܭࢉ࣌ؒ • 3.1GHz ΫΞουίΞ Intel Core i7 Λར༻͢ΔධՁ؀ڥʹ͓͍ͯɼ΢Οϯυ ΢ ͋ͨΓͷܭࢉ࣌ؒ͸1.71ϛϦඵʙ1.75ϛϦඵ

    • ΢ΣϒαΠτͷಡΈࠐΈ࣌ؒ͸1,000ϛϦඵະຬ͕๬·͍͠ͱ͞Ε͓ͯΓɼఏ Ҋख๏ʹΑΔมԽݕग़ʹֻ͔Δ࣌ؒ͸े෼ʹখ͍͞ W′ u (t) 31
  16. ࠓޙʹ͍ͭͯ • ఏҊख๏ͷਫ਼౓ͷվળ • ಛ௃ྔͷ஋͕มԽ͢Δࡍͷਖ਼ෛํ޲ͷϞσϧ΁ͷ૊ΈࠐΈ • ಛ௃ྔͷ஋ͷมಈ͕େ͖͍ظؒͷআ֎ͳͲ • ܭࢉ࣌ؒͷ୹ॖ •

    มԽݕग़ʹ༻͍Δ΢Οϯυ΢Λ֤ཁૉͰ෼ׂͤͣҰՕॴͰ౳෼ׂ͢Δ • খඪຊʹରͯ͠ؤ݈ͳ౷ܭతԾઆݕఆͷख๏ͷݕ౼ 34