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

会社訪問アプリ「Wantedly Visit」における 人・仕事・会社の関係性を考慮した相互推薦システム

Shuhei Goda
March 06, 2023

会社訪問アプリ「Wantedly Visit」における 人・仕事・会社の関係性を考慮した相互推薦システム

2023年3月6日 DEIM2023 (https://event.dbsj.org/deim2023/) における技術報告の資料です。

3b-5: ドメイン指向推薦:企業・人材
3月6日 13:30 ~ 15:30
https://deim-management-system.github.io/deim2023_program/index.html#3b-5

既存研究ではあまり取り扱われていない求職者の会社に関する嗜好の重要性に着目し、人・仕事・会社の3要素に基づいた相互推薦システムの提案手法の性能を会社訪問アプリ「Wantedly Visit」の実データを用いて検証しています。

Shuhei Goda

March 06, 2023
Tweet

More Decks by Shuhei Goda

Other Decks in Research

Transcript

  1. ©2023 Wantedly, Inc.

    ձࣾ๚໰ΞϓϦʮWantedly Visitʯʹ͓͚Δ

    ਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ


    DEIM2023 3b-5: υϝΠϯࢦ޲ਪનɿاۀɾਓࡐʲٕज़ใࠂʳ
    Mar. 6 2023 - ߹ాपฏʢ΢ΥϯςουϦʔגࣜձࣾʣ

    View full-size slide

  2. ©2023 Wantedly, Inc.
    1. اۀ঺հ


    • ձࣾͱϓϩμΫτͷ঺հ


    • ϓϩμΫτʹ͓͚ΔσʔλαΠΤϯεͷ׆༻ࣄྫ


    • ΞΧσϛΞʹ͓͚Δ׆ಈ


    2. ૬ޓਪનγεςϜͱ͸


    • ૬ޓਪનγεςϜͷ֓ཁ


    • HRྖҬʹ͓͚Δઌߦݚڀ


    3. ਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ


    • ໰୊ఏى


    • ఏҊख๏


    • ࣮ݧɾߟ࡯
    ձࣾ๚໰ΞϓϦʮWantedly Visitʯʹ͓͚Δਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ

    View full-size slide

  3. ©2023 Wantedly, Inc.
    1. اۀ঺հ


    • ձࣾͱϓϩμΫτͷ঺հ


    • ϓϩμΫτʹ͓͚ΔσʔλαΠΤϯεͷ׆༻ࣄྫ


    • ΞΧσϛΞʹ͓͚Δ׆ಈ


    2. ૬ޓਪનγεςϜͱ͸


    • ૬ޓਪનγεςϜͷ֓ཁ


    • HRྖҬʹ͓͚Δઌߦݚڀ


    3. ਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ


    • ໰୊ఏى


    • ఏҊख๏


    • ࣮ݧɾߟ࡯
    ձࣾ๚໰ΞϓϦʮWantedly Visitʯʹ͓͚Δਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ

    View full-size slide

  4. ©2023 Wantedly, Inc.
    ࣗݾ঺հ
    ໊લ

    ߹ా पฏ(Shuhei Goda)


    ܦྺ

    2016೥3݄ ๺ւಓେֶେֶӃཧֶӃӉ஦ཧֶઐ߈म࢜՝ఔमྃ


    ॴଐɾ໾ׂ

    ΢ΥϯςουϦʔגࣜձࣾɾData Science Tech Lead


    ձࣾ๚໰ΞϓϦ Wantedly Visit ͷϚονϯά෦෼Λ୲͏

    ਪનγεςϜͷ։ൃΛਪਐ͍ͯ͠ΔɻKaggle Masterɻ @hakubishin3
    @jy_msc

    View full-size slide

  5. ©2023 Wantedly, Inc.
    Create A World Where Work Drives Passion.
    γΰτͰίίϩΦυϧͻͱΛ;΍͢
    ձࣾ঺հ ࢲͨͪͷϛογϣϯ
    ©2023 Wantedly, Inc.

    View full-size slide

  6. ©2023 Wantedly, Inc.
    ձࣾʹ༡ͼʹߦ͜͏
    ňΧδϡΞϧ໘ஊʯͱ͍͏࢓ࣄ୳͠ͷ৽ج४
    • ਓͱνʔϜΛʮ૝͍ʯͰϚονϯά


    • ň࿩Λฉ͖ʹߦ͘ʯମݧΛ௨ͯ͡ϑϥοτͳग़ձ͍Λ૑ग़
    ਐԽΛଓ͚ΔUIͱج൫
    • ίϯςϯπ͕ө͑ΔϞμϯͰγϯϓϧͳUIදݱ


    • ΑΓຊ࣭తͳϚονϯάΛՄೳʹ͢ΔਪનΞϧΰϦζϜ
    ఏڙϓϩμΫτ

    View full-size slide

  7. ©2023 Wantedly, Inc.
    ࣄྫ঺հ ਓͱձࣾͷʮ྆૝͍ʯΛ࣮ݱ͢Δ૬ޓਪનγεςϜ

    View full-size slide

  8. ©2023 Wantedly, Inc.
    ࠃࡍֶձ΁ͷௌߨࢀՃ / Ξ΢τϓοτ

    View full-size slide

  9. ©2023 Wantedly, Inc.
    ࠃ಺ֶձ΁ͷڠࢍͱൃද
    https://www.ai-gakkai.or.jp/jsai2022/industrial-session/
    https://www.ai-gakkai.or.jp/jsai2022/sponsor/
    JSAI
    DEIM
    https://deim-management-system.github.io/deim2023_program/index.html#3b-5
    https://event.dbsj.org/deim2023/

    View full-size slide

  10. ©2023 Wantedly, Inc.
    1. اۀ঺հ


    • ձࣾͱϓϩμΫτͷ঺հ


    • ϓϩμΫτʹ͓͚ΔσʔλαΠΤϯεͷ׆༻ࣄྫ


    • ΞΧσϛΞʹ͓͚Δ׆ಈ


    2. ૬ޓਪનγεςϜͱ͸


    • ૬ޓਪનγεςϜͷ֓ཁ


    • HRྖҬʹ͓͚Δઌߦݚڀ


    3. ਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ


    • ໰୊ఏى


    • ఏҊख๏


    • ࣮ݧɾߟ࡯
    ձࣾ๚໰ΞϓϦʮWantedly Visitʯʹ͓͚Δਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ

    View full-size slide

  11. ©2023 Wantedly, Inc.
    ૬ޓਪનγεςϜ(Reciprocal Recommeder Systems)ͱ͸ʁ
    User Item
    ैདྷͷҰൠతͳਪનγεςϜ
    ૬ޓਪનγεςϜ
    e.g. Amazon, Netflix
    ਪન͞ΕΔ
    ʮαʔϏε಺ͷϢʔβ͕ޓ͍ʹਪન͞Ε߹͏γεςϜʯ
    User(Job Seeker) User(Recruiter/Company)
    e.g. Wantedly, LinkedIn
    ਪન͞Ε߹͏
    User(Female)
    User(Male)
    e.g. Tinder, Pairs
    ਪન͞Ε߹͏

    View full-size slide

  12. ©2023 Wantedly, Inc.
    ૬ޓਪનγεςϜʹ͓͚Δਪનͷ”੒ޭ”ͱ͸ʁ
    User Item
    ैདྷͷҰൠతͳਪનγεςϜ
    ૬ޓਪનγεςϜ
    ޓ͍ͷᅂ޷͕Ұகͯ͠ॳΊͯਪન͕"੒ޭ"ͨ͜͠ͱʹͳΔ


    Like
    Like
    Nope
    Like
    Like
    User
    User
    User
    User

    View full-size slide

  13. ©2023 Wantedly, Inc.
    ૬ޓਪનγεςϜͷ࿮૊Έ
    1. γεςϜ಺ͷϢʔβ͸ A ͱ B ͷ̎ͭͷάϧʔϓʹ෼͔Ε͓ͯΓɺҟͳΔάϧʔϓͷϢʔβ͕ޓ͍ਪન͞ΕΔ

    ΋ͷͱ͢Δʢe.g. σʔςΟϯάαʔϏεʹ͓͚ΔஉঁɺٻਓαʔϏεʹ͓͚Δٻ৬ऀͱاۀʣ


    2. ୯ํ޲ͷᅂ޷ͷେ͖͞Λද͢ Preference Score ΛɺA ͔Β B ΁ͷϢʔβٴͼ B ͔Β A ͷϢʔβͷͦΕͧΕ

    ʹ͍ͭͯܭࢉ


    3. Aggregation Function Λར༻ͯ͠ɺ̎ͭͷ Preference Score Λ૊Έ߹Θͤͯ૒ํ޲ͷᅂ޷ͷେ͖͞Λද͢

    Reciprocal Preference Score Λܭࢉ
    Predict User A to User B


    Preference Score
    Predict User B to User A


    Preference Score
    Aggregate

    Preference Scores
    Reciprocal

    Preference Score

    View full-size slide

  14. ©2023 Wantedly, Inc.
    HRྖҬʹ͓͚Δ૬ޓਪનγεςϜͷઌߦݚڀ
    ͜Ε·Ͱʮਓ(ީิऀ)ʯͱʮ࢓ࣄ(δϣϒɾσΟεΫϦϓγϣϯ)ʯͷཧ૝తͳϚονϯάΛ͍͔ʹ࣮ݱ͢Δ͔ɺ

    ʹয఺Λ౰ͯͯɺHRྖҬʹ͓͚Δ૬ޓਪનγεςϜͷݚڀ͕ਐΜͰ͍Δ
    Tsunenori 2013 Yang 2022
    Job
    User
    ʁ
    ɾ৬छ


    ɾ࢓ࣄ಺༰


    ɾۈ຿஍


    ɾۈ຿࣌ؒ


    ɾ࠾༻ܗଶ


    ɾ଴۰


    ɹ….

    View full-size slide

  15. ©2023 Wantedly, Inc.
    1. اۀ঺հ


    • ձࣾͱϓϩμΫτͷ঺հ


    • ϓϩμΫτʹ͓͚ΔσʔλαΠΤϯεͷ׆༻ࣄྫ


    • ΞΧσϛΞʹ͓͚Δ׆ಈ


    2. ૬ޓਪનγεςϜͱ͸


    • ૬ޓਪનγεςϜͷ֓ཁ


    • HRྖҬʹ͓͚Δઌߦݚڀ


    3. ਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ


    • ໰୊ఏى


    • ఏҊख๏


    • ࣮ݧɾߟ࡯
    ձࣾ๚໰ΞϓϦʮWantedly Visitʯʹ͓͚Δਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનγεςϜ

    View full-size slide

  16. ©2023 Wantedly, Inc.
    ਓ͸ԿΛߟྀ͠ͳ͕Β࢓ࣄΛ୳͢ͷ͔
    Ұൠతʹਓ͕࢓ࣄΛ୳࣌͢ʹߟྀ͢Δͷ͸ʮδϣϒɾσΟεΫϦϓγϣϯʯ͚ͩͰͳ͍͸ͣ


    • ྫ͑͹ɺձࣾͷϛογϣϯ΍ձࣾͰಇ͍͍ͯΔϝϯόʔͷงғؾ


    • ձࣾͷن໛΍੡඼ɺۀքɺઓུͳͲ΋ॏཁͩΖ͏
    Job
    User
    ɾ৬छ


    ɾ࢓ࣄ಺༰


    ɾۈ຿஍


    ɾۈ຿࣌ؒ


    ɾ࠾༻ܗଶ


    ɾ଴۰


    ɹ….
    ʮ࢓ࣄʯʹؔ͢Δᅂ޷
    Company
    ɾϛογϣϯ


    ɾҰॹʹಇ͘ਓͨͪ


    ɾձࣾن໛


    ɾۀք


    ɾ੡඼


    ɾઓུ


    ɹ…
    ʮձࣾʯʹؔ͢Δᅂ޷
    Company಺

    View full-size slide

  17. ©2023 Wantedly, Inc.
    طଘݚڀͰߟྀͰ͖͍ͯͳ͍͜ͱ͸Կ͔
    ͜Ε·Ͱʮਓʯͱʮ࢓ࣄʯͷؒͷΠϯλϥΫγϣϯ͔͠ѻΘΕͯ͜ͳ͔ͬͨͷͰɺ

    ʮਓʯͱʮձࣾʯͷؔ܎Λද͢৘ใΛ໌ࣔతʹऔΓೖΕΔ͜ͱ͸Ͱ͖͍ͯͳ͍


    • ࢓ࣄʹؔ͢Δᅂ޷͸நग़Ͱ͖Δ͕ɺձࣾʹؔ͢Δᅂ޷͸ʁ
    Job
    User
    ɾ৬छ


    ɾ࢓ࣄ಺༰


    ɾۈ຿஍


    ɾۈ຿࣌ؒ


    ɾ࠾༻ܗଶ


    ɾ଴۰


    ɹ….
    ʮਓʯͱʮ࢓ࣄʯͷؔ܎ੑ
    Company
    ɾϛογϣϯ


    ɾҰॹʹಇ͘ਓͨͪ


    ɾձࣾن໛


    ɾۀք


    ɾ੡඼


    ɾઓུ


    ɹ…
    ʮਓʯͱʮձࣾʯͷؔ܎ੑ
    Company಺

    View full-size slide

  18. ©2023 Wantedly, Inc.
    ձࣾ๚໰ΞϓϦ Wantedly Visit ͱ͸
    ձࣾ΁ͷڞײΛ࣠ͱͨ͠࢓ࣄ୳͠ͷମݧΛఏڙ͢ΔαʔϏε


    • ࢓ࣄ΋ॏཁͳཁૉ͕ͩɺͦΕҎ্ʹձࣾʹڞײͰ͖Δ͔Λॏཁࢹ


    • ձࣾͷϛογϣϯ΍ͦ͜Ͱಇ͘ਓͷ૝͍͕఻ΘΓ΍͍͢ϓϩμΫτઃܭ
    Ϣʔβ
    ձࣾͷίϯςϯπ


    ɾձࣾϖʔδ


    ɾձࣾϒϩά
    ࢓ࣄͷίϯςϯπ


    ɾืू
    ճ༡
    ืू΁ͷ

    Ԡื
    ձࣾʹ࿩Λฉ͖ʹߦ͘

    View full-size slide

  19. ©2023 Wantedly, Inc.
    Wantedly Visit ʹطଘͷ૬ޓਪનγεςϜΛద༻͢Δࡍͷ՝୊
    • ҰൠతͳδϣϒϚονϯάαʔϏεͱൺ΂ͯɺϢʔβʔମݧʹ͓͍ͯ

    ձࣾͷཁૉ͕ॏཁͳ໾ׂΛ୲͍ͬͯΔ


    • ͦͷͨΊɺʮਓʯͱʮ࢓ࣄʯͷΠϯλϥΫγϣϯ͚ͩΛߟྀͨ͠طଘͷ૬ޓਪનγεςϜ

    ͷ࿮૊ΈΛͦͷ··ద༻͢Δͷ͸ෆे෼ͩͱߟ͑Δ


    -
    ⭕Ϣʔβͷ࢓ࣄʹؔ͢Δᅂ޷ɺ࢓ࣄͷީิऀʹؔ͢Δᅂ޷ʢ৬຿ΛՌͨͤΔظ଴͕͋Δ͔Ͳ͏͔ʣ


    -
    ❌Ϣʔβͷձࣾʹؔ͢Δᅂ޷ɺձࣾͷީิऀʹؔ͢Δᅂ޷ʢՁ஋؍ͷҰகɺWillͱMustͷҰகɺͳͲʣ

    View full-size slide

  20. ©2023 Wantedly, Inc.
    Wantedly Visit ʹదͨ͠૬ޓਪનγεςϜͱ͸
    ʮਓʯͱʮ࢓ࣄʯɺʮਓʯͱʮձࣾʯͷؔ܎ੑΛ໌ࣔతʹऔΓೖΕͨ૬ޓਪનγεςϜͷఏҊ


    • ఏҊख๏Λ Wantedly Visit ͷ࣮σʔλΛར༻ͯ͠ݕূ͢Δ
    Job
    User
    ɾ৬छ


    ɾ࢓ࣄ಺༰


    ɾۈ຿஍


    ɾۈ຿࣌ؒ


    ɾ࠾༻ܗଶ


    ɾ଴۰


    ɹ….
    ࢓ࣄʹؔ͢Δᅂ޷
    Company
    ɾϛογϣϯ


    ɾҰॹʹಇ͘ਓͨͪ


    ɾձࣾن໛


    ɾۀք


    ɾ੡඼


    ɾઓུ


    ɹ…
    ձࣾʹؔ͢Δᅂ޷
    Company಺
    Ϣʔβʹؔ͢Δᅂ޷
    Ϣʔβʹؔ͢Δᅂ޷

    View full-size slide

  21. ©2023 Wantedly, Inc.
    ࣮ݧ֓ཁ
    Wantedly Visit ʹ͓͚ΔϢʔβͱืूͷ Matching Λ༧ଌ (ਪન୯Ґ͸ืूͰ͋ΔͨΊ)


    • Ϣʔβ͕Ԡื or ืू(اۀ)͕εΧ΢τૹ৴ͨ͠ࡍʹ૬ख͕ϝοηʔδΛฦ৴͢Ε͹ Match (positive)


    • Ԡื or εΧ΢τૹ৴Λ૬ख͕Ӿཡ্ͨ͠ͰϝοηʔδΛฦ৴͠ͳ͚Ε͹ Not Match (negative)
    ืू(اۀ)
    Ϣʔβ
    Ԡื
    ϝοηʔδฦ৴
    ืू(اۀ)
    Ϣʔβ
    εΧ΢τૹ৴
    ϝοηʔδฦ৴
    ืू(اۀ)
    Ϣʔβ
    Ԡื
    ӾཡͷΈ
    ืू(اۀ)
    Ϣʔβ
    εΧ΢τૹ৴
    ӾཡͷΈ
    Match Not Match

    View full-size slide

  22. ©2023 Wantedly, Inc.
    ࣮ݧσʔλ
    • Wantedly Visit ʹ͓͚Δ 2022೥04݄ ʙ 2023೥02݄ ͷ໿1೥෼ͷߦಈϩά


    • ֘౰ظؒதʹ5݅Ҏ্ͷᅂ޷σʔλΛ༗͢ΔϢʔβͱืू(اۀ)


    • ֘౰ظؒதʹ100݅Ҏ্ͷᅂ޷σʔλΛ༗͢ΔϢʔβΛআ֎

    View full-size slide

  23. ©2023 Wantedly, Inc.
    ධՁํ๏
    • ධՁ஋͕ෆ໌Ͱͳ͍Ϣʔβͱืूͷ૊Έ߹Θͤͷ͏ͪ 10% Λςετσʔλͱͯ͠ར༻


    • ༧ଌ͞Εͨ Reciprocal Preference Score Λ AUC ͰධՁ

    View full-size slide

  24. ©2023 Wantedly, Inc.
    ϕʔεϥΠϯख๏
    Preference Scoreͷਪ࿦ʹߦྻ෼ղΛ༻͍ͯɺू໿ؔ਺͸ௐ࿨ฏۉΛ࠾༻(Neve 2019)


    • طଘݚڀʹ͓͚ΔHRͷ૬ޓਪનͰ͸ɺϢʔβͱ࢓ࣄͷΠϯλϥΫγϣϯΛߦྻ෼ղ΍GNNͰѻ͍ͬͯΔ


    • ݕূͷ୯७ԽͷͨΊʹɺࠓճ͸ߦྻ෼ղΛબ୒
    Predict User to Project


    Preference Score
    Predict Project to User


    Preference Score
    Aggregate using

    Harmonic Mean
    Ϣʔβ × ืू ͷ

    ΠϯλϥΫγϣϯߦྻ

    View full-size slide

  25. ©2023 Wantedly, Inc.
    ఏҊख๏
    2छྨͷΠϯλϥΫγϣϯߦྻΛ༻͍ͯ૬ޓਪનείΞΛਪ࿦͢Δ


    • ϢʔβͱձࣾͷؒͷΠϯλϥΫγϣϯ͔Βɺձࣾʹؔ͢ΔϢʔβͷ܏޲Λநग़


    • ձࣾʹؔ͢Δᅂ޷ͱ࢓ࣄʹؔ͢Δᅂ޷Λू໿(ௐ࿨ฏۉ or ࢉज़ฏۉ)͔ͯ͠Βɺ૬ޓਪનείΞΛࢉग़͢Δ
    Predict User to Company


    Preference Score
    Predict Company to User


    Preference Score
    Aggregate using

    HM / AM
    Predict User to Project


    Preference Score
    Predict Project to User


    Preference Score
    Aggregate using

    HM / AM
    Ϣʔβ × ืू ͷ

    ΠϯλϥΫγϣϯߦྻ
    Ϣʔβ × ձࣾͷ

    ΠϯλϥΫγϣϯߦྻ
    Aggregate using

    Harmonic Mean

    View full-size slide

  26. ©2023 Wantedly, Inc.
    ࣮ݧ݁Ռᶃ
    ϕʔεϥΠϯʹൺ΂ͯɺఏҊख๏ʹΑΔ Reciprocal Preference Score ͷ༧ଌੑೳ͸޲্


    • ௐ࿨ฏۉΑΓ΋ࢉज़ฏۉͷ΄͏͕ߴ͍ੑೳʹɺ͔͠͠ू໿ؔ਺ͷҧ͍ʹΑΔࠩ͸େ͖͘ͳ͍


    • େ͖ͳ޲্ͷ༨஍͸͋Γͦ͏ɺ཈͑Δ΂͖ϙΠϯτ͕͋ΔͷͰ͸ʁ
    AUC Reciprocal Preference Score Rate of change (%)
    ϕʔεϥΠϯ 0.7122 -
    ఏҊख๏(ௐ࿨ฏۉ) 0.7269 +2.08%
    ఏҊख๏(ࢉज़ฏۉ) 0.7301 +2.52%

    View full-size slide

  27. ©2023 Wantedly, Inc.
    Ծઆग़͠ͱͦΕʹجͮ͘ఏҊख๏ͷमਖ਼
    ϢʔβʹΑͬͯ࢓ࣄͷᅂ޷ͱձࣾͷᅂ޷ͷॏΈͷڧ͕͞ҟͳΔͷͰ͸ʁ


    • ྫ͑͹ɺస৬ҙཉͷߴ͍Ϣʔβ΄Ͳձࣾʹؔ͢ΔཁૉΛॏཁࢹ͢Δ͔΋͠Εͳ͍


    • Ϣʔβͷᅂ޷ͷू໿ؔ਺ʹϩδεςΟοΫճؼΛ࠾༻͠ɺ࠷దͳॏΈͰू໿Ͱ͖ΔΑ͏ʹ͢Δ
    Predict User to Company


    Preference Score
    Predict Company to User


    Preference Score
    Aggregate using

    HM / AM
    Predict User to Project


    Preference Score
    Predict Project to User


    Preference Score
    Aggregate using

    Logistic Reg
    Ϣʔβ × ืू ͷ

    ΠϯλϥΫγϣϯߦྻ
    Ϣʔβ × ձࣾͷ

    ΠϯλϥΫγϣϯߦྻ
    Aggregate using

    Harmonic Mean

    View full-size slide

  28. ©2023 Wantedly, Inc.
    ࣮ݧ݁Ռᶄ
    Ϣʔβ͕ઃఆͨ͠స৬ҙཉΛ༻͍ͯɺϢʔβͷᅂ޷ΛϩδεςΟοΫճؼͰू໿ͤ͞Δ


    • ͜ͷ΍Γํ͸ޮՌͳ͠
    AUC Reciprocal Preference Score Rate of change (%)
    ϕʔεϥΠϯ 0.7122 -
    ఏҊख๏(ௐ࿨ฏۉ) 0.7269 +2.08%
    ఏҊख๏(ࢉज़ฏۉ) 0.7301 +2.52%
    ఏҊख๏(ϩδճؼ) 0.7255 +2.07%

    View full-size slide

  29. ©2023 Wantedly, Inc.
    ߟ࡯
    ͳͥఏҊख๏Ͱ͸૝ఆ΄Ͳͷ༧ଌਫ਼౓ͷେ͖ͳ޲্͕ݟࠐΊͳ͔ͬͨͷ͔


    1. ϕʔεϥΠϯͷ࣌఺Ͱ͢ͰʹɺձࣾͷཁૉΛ͋Δఔ౓நग़Ͱ͖͍ͯΔʁ


    2. ఏҊख๏Ͱ͸ɺձࣾͷཁૉΛ͏·͘நग़Ͱ͖͍ͯͳ͍ʁ


    3. Ϣʔβ͸ձࣾͷཁૉͰ࢓ࣄΛ୳͢͜ͱ͸ͳ͍ʁ


    - ఆੑతʹ͸͋Γಘͳ͍ͷͰ͜ͷઢ͸٫ԼʢϢʔβώΞϦϯάɺ͜Ε·Ͱͷ෼ੳͷ݁ՌͳͲʣ

    View full-size slide

  30. ©2023 Wantedly, Inc.
    ߟ࡯ᶃ
    ϕʔεϥΠϯͷ࣌఺Ͱ͢ͰʹɺձࣾͷཁૉΛ͋Δఔ౓நग़Ͱ͖͍ͯΔͷͰ͸ͳ͍͔


    • Wantedly Visit ʹ͓͚Δืूϖʔδͷߏ੒͸ҎԼͷ௨Γɺձࣾͷཁૉͷํ͕༏ઌ౓ߴ͍


    • Ϣʔβ͸͜ͷߏ੒ʹڧ͘Өڹ͞ΕɺืूͱϢʔβͷΠϯλϥΫγϣϯߦྻʹ΋ձࣾͷཁૉؚ͕·ΕΔͷͰ͸
    Χόʔࣸਅ
    ձࣾͷϝϯόʔ঺հ
    ձࣾͷ8)"5 ࣄۀ಺༰ͷઆ໌ͳͲ

    ձࣾͷ8): ϏδϣϯɺϛογϣϯͳͲ

    ձࣾͷ)08 ࣄۀͷಛ௃ɺಇ͘஥ؒͳͲ

    ϙδγϣϯͷৄࡉɺۀ຿಺༰ͳͲ
    ืूͷλΠτϧɺ৬छɺ࠾༻ܗଶɺձ໊ࣾ
    ืूϖʔδͷߏ੒
    ϖʔδ্෦
    ϖʔδԼ෦
    ձࣾͷཁૉ
    ࢓ࣄͷཁૉ
    ྆ํͷཁૉ

    View full-size slide

  31. ©2023 Wantedly, Inc.
    ߟ࡯ᶄ
    ఏҊख๏Ͱ͸ɺձࣾͷཁૉΛ͏·͘நग़Ͱ͖͍ͯͳ͍ͷͰ͸ͳ͍͔


    • Wantedly Visit ʹ͓͍ͯϢʔβ͸࢓ࣄʹؔ͢Δ࠷௿৚݅ʢޏ༻ܗଶɺۈ຿஍ɺ৬छͳͲʣΛ

    ઃఆͯ͠ϑΟϧλϦϯάΛ͔ͯ͠ΒձࣾΛ୳͢


    • ྫ͑͹ɺձࣾͷཁૉ͕ຬͨ͞Ε͍ͯͯ΋ɺ΍Γ͍ͨ࢓ࣄ͕ͳ͍ͳΒԠื͠ͳ͍ͩΖ͏


    • ձࣾ - ϢʔβͷΠϯλϥΫγϣϯʹ͸ɺ࢓ࣄͷཁૉ΋େؚ͖͘·Ε͍ͯΔͷͰ͸ͳ͍͔
    Wantedly Visit ʹ͓͚Δืूͷݕࡧ৚݅


    ޏ༻ܗଶɺۈ຿஍ɺ৬छͳͲͷ৚݅ઃఆ͕Ͱ͖Δ

    View full-size slide

  32. ©2023 Wantedly, Inc.
    Future work
    Wantedly Visit ʹదͨ͠ɺਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનͷߏ੒͸ʁ


    • ࢓ࣄ - ਓͷΠϯλϥΫγϣϯཤྺΛϕʔεʹɺձࣾͷ৘ใΛ෇༩͢ΔʢFactorization MachineͳͲʣ


    • ͜ͷߏ੒ͷ৔߹ɺձࣾͷಛ௃நग़Λ໌ࣔతʹߦ͏ඞཁ͕͋Δʢϛογϣϯ΍ձࣾͷงғؾɺ

    ঎඼΍ۀքɺձࣾن໛΍ઓུͳͲʣ


    - Wantedly Visit ʹ৘ใ͕ଘࡏ͠ͳ͍΋ͷ͸ɺ৘ใΛऔಘͰ͖ΔΑ͏ʹ͢Δମ੍ͷߏங


    - Wantedly Visit ʹ৘ใ͸ଘࡏ͍ͯ͠Δ͕ඇߏ଄ԽσʔλʢςΩετɺը૾ʣͱͯ͋͠Δ΋ͷ͸ɺ໌ࣔతͳಛ௃

    ͱͯ͠ѻ͑ΔΑ͏ʹ৘ใநग़͕ඞཁ


    - ಛʹϛογϣϯ΍ձࣾͷงғؾͳͲͷந৅తͳཁૉ͸ɺը૾৘ใ΍ςΩετ͔Β͏·͘நग़͢Δ࿮૊Έ͕ඞཁ
    Wantedly Visit ʹ͓͚Δืूը૾ͷऔΓ૊Έ͸

    DEIM2022ͷٕज़ใࠂͰൃද

    View full-size slide

  33. ©2023 Wantedly, Inc.
    ݁࿦
    ໨త: Wantedly Visit ʹదͨ͠ɺਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛߟྀͨ͠૬ޓਪનͷߏ੒ݕ౼


    Ϣʔβͷ࢓ࣄͱձࣾͷᅂ޷ਪఆΛ੾Γ෼͚ͯɺ࢓ࣄͷᅂ޷ͱձࣾͷᅂ޷ΛϢʔβ͝ͱʹՃॏฏۉ͢Δ

    ͜ͱΛఏҊɺେ͖ͳվળ͸ݟࠐΊͳ͔ͬͨ


    ߟ࡯: ձࣾͱ࢓ࣄͷᅂ޷Λ͏·͘੾Γ෼͚ΒΕ͍ͯͳ͍Մೳੑ͕͋Δ


    1. αʔϏεʹ͓͚Δืूίϯςϯπͷಛ௃: ձࣾͷཁૉ͕ଟؚ͘·Ε͍ͯΔ


    2. αʔϏεʹ͓͚ΔϢʔβߦಈͷಛ௃: ࢓ࣄʹؔ͢Δ࠷௿৚݅Λઃఆ͔ͯ͠ΒձࣾΛ୳͢


    ࠓޙ: ਓɾ࢓ࣄɾձࣾͷؔ܎ੑΛ׆͔͠੾ΔͨΊʹɺͲ͏͍͏ߏ੒͕ద੾ͳͷ͔


    • ࢓ࣄ - ਓͷΠϯλϥΫγϣϯཤྺΛϕʔεʹɺձࣾͷ৘ใΛ෇͚Ճ͑Δ


    • ը૾΍ςΩετͳͲͷඇߏ଄Խσʔλ͔Βɺཉ͍͠ձࣾͷಛ௃Λ͏·͘৘ใநग़͢Δඞཁ͕͋Δ

    View full-size slide

  34. ©2023 Wantedly, Inc.
    3FGFSFODFT
    • (Pizzato 2010) Luiz Pizzato, Tomek Rej, Thomas Chung, Irena Koprinska, and Judy Kay. 2010. RECON: a reciprocal recommender for online
    dating. Proceedings of the fourth ACM conference on Recommender systems, 207-214.


    • (Hongtao 2011)Hongtao Yu, C. Liu and Fuzhi Zhang. 2011. Reciprocal Recommendation Algorithm for the Field of Recruitment. Journal of
    Information & Computational Science, 4061-4068.


    • (Tsunenori 2013)Tsunenori Mine, Tomoyuki Kakuta, Akira Ono. 2013. Reciprocal Recommendation for Job Matching with Bidirectional Feedback.
    2013 Second IIAI International Conference on Advanced Applied Informatics, 39-44.


    • (Neve 2019)J Neve, I Palomares.2019. Latent Factor Models and Aggregation Operators for Collaborative Filtering in Reciprocal Recommender
    Systems Proceedings of the 13th ACM Conference on Recommender Systems, 219-227.


    • (Yang 2022)Yang, Chen and Hou, Yupeng and Song, Yang and Zhang, Tao and Wen, Ji-Rong and Zhao and Wayne Xin. 2022. Modeling Two-Way
    Selection Preference for Person-Job Fit. Proceedings of the 16th ACM Conference on Recommender Systems,102-112.

    View full-size slide