Tweet
Tweet

Slide 1

Slide 1 text

"*γεςϜ։ൃͷϥΠϑαΠΫϧʹ ਓؒͷೝ஌όΠΞε͕༩͑ΔӨڹ ʙ"*Λͭ͘Δɾ͔ͭ͏ਓؒΛ஌Δʙ ஜ೾େֶҩֶҩྍܥ ߨࢣ ߳઒ཨಸ 4QFDJBMUIBOLTUPɿখྛ࢙໌͞Μʢ໌࣏େֶʣɺؔྙٱ͞Μʢ౦ژେֶʣɺന࠭େ͞Μʢ௥ख໳ֶӃେֶʣ ຊൃද͸ɺҎԼͷࢧԉΛड͚͍ͯ·͢ɻ +45͖͕͚͞ʢ+1.+13*ʣɺ+45ະདྷࣾձ૑଄ࣄۀʢ+1.+.*(ʣɺ/&%0ʢ+1/1ʣɺՊݚඅʢ,ʣ ࢈૯ݚ ਓ޻஌ೳηϛφʔ

Slide 2

Slide 2 text

/ 86 Take Home Message 00 ਺஋ʹͨ͠ॠؒʹҰਓา͖ͯ͠͠·͏΋ͷ͸ੈʹଟ͋͘Δɻ ͋͋ݴ͑͹ɺ͜͏ݴ͏ɻ

Slide 3

Slide 3 text

/ 86 AIʹΑΔਓؒͷҙࢥܾఆͷࢧԉ 02 AI͸ਓؒͷ஌ੑΛɾɾɾ ௒ӽ͢Δ ࠶ݱ͢Δ ࢧԉɾ֦ு͢Δಓ۩ ൚༻ਓ޻஌ೳ ௒஌ೳ γϯΪϡϥϦςΟ "*ΞϥΠϝϯτ ҙࢥܾఆࢧԉ ߹ҙܗ੒ࢧԉ

Slide 4

Slide 4 text

/ 86 AIʹΑΔਓؒͷҙࢥܾఆͷࢧԉ 02 AI͸ਓؒͷ஌ੑΛɾɾɾ ௒ӽ͢Δ ࠶ݱ͢Δ ࢧԉɾ֦ு͢Δಓ۩ uࣾձ͕ɺҙࢥܾఆऀͱͯ͠ͷ ੹೚ΛਓؒʹٻΊΔ෼໺ └ҩྍ਍அɺࡋ൑ɺ اۀਓࣄɺߦ੓ͳͲ uਓ͕ؒAIͷग़ྗʹج͍ͮͯ ೳಈతʹҙࢥܾఆ͢Δ෼໺ └޿ࠂɺαʔϏεఏڙͳͲ ൚༻ਓ޻஌ೳ ௒஌ೳ γϯΪϡϥϦςΟ "*ΞϥΠϝϯτ ҙࢥܾఆࢧԉ ߹ҙܗ੒ࢧԉ

Slide 5

Slide 5 text

/ 86 AIʹΑΔਓؒͷҙࢥܾఆͷࢧԉ 02 [1] ྩ࿨̐೥౓਍ྍใुվఆʹ͍ͭͯʢ௨஌ʣ, อൃ 0304 ୈ 1 ߸, 2022 [2] Brennan, T., Dieterich, W., & Ehret, B. (2009). Evaluating the predictive validity of the COMPAS risk and needs assessment system. Criminal Justice and behavior, 36(1), 21-40. [3] https://www.darpa.mil/program/explainable-artificial-intelligence [4] https://www.robustintelligence.com/jp-whitepaper/ai- governance-whitepaper-1 [5] https://www.digiarc.aist.go.jp/publication/aiqm/ [13] Vasey, B., Nagendran, M., Campbell, B., Clifton, D. A., Collins, G. S., Denaxas, S., ... & McCulloch, P. (2022). Reporting guideline for the early-stage clinical evaluation of decision support systems driven by artificial intelligence: DECIDE-AI. Nature medicine, 28(5), 924-933. [14] https://pubs.rsna.org/page/ai/blog/2022/09/ryai_editorsblog0928 [6] https://www.soumu.go.jp/menu_news/s-news/01tsushin06_02000277.html [7] https://www8.cao.go.jp/cstp/aigensoku.pdf [8] https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines- trustworthy-ai [9] https://standards.ieee.org/wp-content/uploads/import/documents/other/ead_v2.pdf [10] https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449 ࿨༁ɿ https://www.soumu.go.jp/main_content/000642217.pdf, 2019 [11] https://unesdoc.unesco.org/ark:/48223/pf0000385082 [12] https://www.un.org/en/ai-advisory-body ظ଴ͷߴ·Γɾ࣮Ԡ༻ͷ޿͕Γ ݒ೦΁ͷରԠ u ෯޿͍ର৅ ◇ ҩྍ ը૾਍அ"* ʙຊ๜อݥద༻ <> ◇ ࢘๏ $0.1"4<> ◇ ܉ࣄ %"31"9"*ϓϩδΣΫτ <> u "*ͷ඼࣭อূ ◇"*ΨόφϯεϗϫΠτϖʔύʔWFS <> ◇ػցֶश඼࣭ϚωδϝϯτΨΠυϥΠϯ <> u ࿦จࣥචͷΨΠυϥΠϯ ◇%&$*%&"* ͳͲ < > u (޿ౡ"*ϓϩηε (σδλϧɾٕज़ֳ྅੠໌ <> u "*ࣾձݪଇͳͲ ◇೔ຊ੓෎ʮਓؒத৺ͷ"*ࣾձݪଇʯ<> ݪଇΛ೥಺ܾఆํ਑ʁ ◇&6ʮ&UIJDT (VJEFMJOFTGPSUSVTUXPSUIZ"*ʯ<> ◇*&&&ʮ&UIJDBMMZ "MJHOFE%FTJHOʯ<> ◇0&$%ʮ3FDPNNFOEBUJPOPGUIF$PVODJMPO "SUJGJDJBM*OUFMMJHFODFʯ<> ◇6&4$0ʮ6/&4$0`T3FDPNNFOEBUJPOPO UIF&UIJDT PG"SUJGJDJBM*OUFMMJHFODFLFZ GBDUTʯ<> u ͦͷଞɺࠃࡍతͳ૊৫ ◇ࠃ࿈ʮ)JHIMFWFM"EWJTPSZ#PEZPO"SUJGJDJBM *OUFMMJHFODFʯ<> ࣾձʹ͓͚Δ"*ͷՁ஋ΛߴΊͯɺ"*͕ࣾձͷՁ஋ΛΑΓߴΊΔͨΊʹ "*͸ຊ౰ʹ૝ఆ௨Γʹ࡞ΒΕͯ࢖ΘΕ͍ͯΔͷ͔ʁຊ౰ʹ࢖ΘΕΔ"*ͱ͸Կ͔ʁ

Slide 6

Slide 6 text

/ 86 AI։ൃϥΠϑαΠΫϧͷ֤ՕॴͰੜ͡ΔόΠΞε 02 Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). ֶशɾਪ࿦ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ

Slide 7

Slide 7 text

/ 86 AI։ൃϥΠϑαΠΫϧͷ֤ՕॴͰੜ͡ΔόΠΞε 02 Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ

Slide 8

Slide 8 text

/ 86 AI։ൃϥΠϑαΠΫϧͷ֤ՕॴͰੜ͡ΔόΠΞε 02 Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ

Slide 9

Slide 9 text

/ 86 AI։ൃϥΠϑαΠΫϧͷ֤ՕॴͰੜ͡ΔόΠΞε 02 Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ

Slide 10

Slide 10 text

/ 86 AI։ൃϥΠϑαΠΫϧͷ֤ՕॴͰੜ͡ΔόΠΞε 02 Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ

Slide 11

Slide 11 text

/ 86 AI։ൃϥΠϑαΠΫϧͷ֤ՕॴͰੜ͡ΔόΠΞε 02 Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ ྺ࢙తͳภΓ ɾஉঁͰҟͳΔ ऩೖσʔλ͕஝ੵ ୅දͷภΓ ɾ೥ྸ΍ਓछͷภΓ ܭଌͷภΓ ɾܭଌޡࠩ ɾσʔλͷඪ४Խෆ଍ ධՁͷภΓ ɾෆద੾ͳධՁࢦඪ ɾϕϯνϚʔΫσʔλ͕ ฼ूஂΛ୅ද͍ͯ͠ͳ͍ ࣮૷ͷภΓ ɾ࠶൜༧ଌϞσϧΛܐظܾఆ γεςϜʹ૊ΈࠐΉ ɾར༻ऀͷࣗ཯ੑΛଛͳ͏ઃܭ

Slide 12

Slide 12 text

/ 86 AI։ൃϥΠϑαΠΫϧͱਓؒͷೝ஌όΠΞε 02 Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ ਓؒʹΑΔภΓ ೝ஌όΠΞε ਓؒʹΑΔภΓ ೝ஌όΠΞε ਓؒʹΑΔภΓ ೝ஌όΠΞε ྺ࢙తͳภΓ ɾஉঁͰҟͳΔ ऩೖσʔλ͕஝ੵ ୅දͷภΓ ɾ೥ྸ΍ਓछͷภΓ ܭଌͷภΓ ɾܭଌޡࠩ ɾσʔλͷඪ४Խෆ଍ ධՁͷภΓ ɾෆద੾ͳධՁࢦඪ ɾϕϯνϚʔΫσʔλ͕ ฼ूஂΛ୅ද͍ͯ͠ͳ͍ ࣮૷ͷภΓ ɾ࠶൜༧ଌϞσϧΛܐظܾఆ γεςϜʹ૊ΈࠐΉ ɾར༻ऀͷࣗ཯ੑΛଛͳ͏ઃܭ

Slide 13

Slide 13 text

/ 86 AI։ൃϥΠϑαΠΫϧͱਓؒͷೝ஌όΠΞε 02 ਆͷΈͧ஌Δ ਅͷੈք Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ ྺ࢙తͳภΓ ɾஉঁͰҟͳΔ ऩೖσʔλ͕஝ੵ ୅දͷภΓ ɾ೥ྸ΍ਓछͷภΓ ܭଌͷภΓ ɾܭଌޡࠩ ɾσʔλͷඪ४Խෆ଍ ධՁͷภΓ ɾෆద੾ͳධՁࢦඪ ɾϕϯνϚʔΫσʔλ͕ ฼ूஂΛ୅ද͍ͯ͠ͳ͍ ਓؒͷೝ஌ͷϑΟϧλʔΛհ͞ͳ͍ͱ֎ࡏԽͰ͖ͳ͍஌ ओ؍త൑அɺଟ͘ͷ࣬ױͷ਍அɺͳͲ ࣮૷ͷภΓ ɾ࠶൜༧ଌϞσϧΛܐظܾఆ γεςϜʹ૊ΈࠐΉ ɾར༻ऀͷࣗ཯ੑΛଛͳ͏ઃܭ ਓؒʹΑΔภΓ ೝ஌όΠΞε ਓؒʹΑΔภΓ ೝ஌όΠΞε ਓؒʹΑΔภΓ ೝ஌όΠΞε

Slide 14

Slide 14 text

/ 86 ࠓ೔ͷ͓͸ͳ͠ 02 AI։ൃϥΠϑαΠΫϧʹ͓͍ͯɺਓؒͷೝ஌όΠΞε͕༩͑ΔӨڹΛ֓؍͢Δ u ਓؒͷҙࢥܾఆ͸ภΔɺͱ͍͏ࣄ࣮Λ஌Δ u ࠓ͋ΔղܾࡦͱɺࠓޙඞཁͳղܾࡦΛߟ͑Δ └पล෼໺ʢHCIͳͲʣ΍ਓؒཧղʢೝ஌ՊֶͳͲʣͷ؍఺΋ ࠓ೔ͷ໨త Schwartz, R., Vassilev, A., Greene, K., Perine, L., Burt, A., & Hall, P. (2022). Towards a standard for identifying and managing bias in artificial intelligence. NIST special publication, 1270(10.6028). ਓؒͷߴ࣍ೝ஌ಛੑʹىҼ͢Δ AIͷόΠΞε͸ݟա͝͞Ε͖ͯͨ

Slide 15

Slide 15 text

/ 86 ࠓ೔ͷ͓͸ͳ͠ 02 AI։ൃϥΠϑαΠΫϧʹ͓͍ͯɺਓؒͷೝ஌όΠΞε͕༩͑ΔӨڹΛ֓؍͢Δ u ਓؒͷҙࢥܾఆ͸ภΔɺͱ͍͏ࣄ࣮Λ஌Δ u ࠓ͋ΔղܾࡦͱɺࠓޙඞཁͳղܾࡦΛߟ͑Δ └पล෼໺ʢHCIͳͲʣ΍ਓؒཧղʢೝ஌ՊֶͳͲʣͷ؍఺΋ ࠓ೔ͷ໨త Birhane, A., Kalluri, P., Card, D., Agnew, W., Dotan, R., & Bao, M. (2022, June). The values encoded in machine learning research. In Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency (pp. 173-184). NeurIPSͱICML 2008/9೥→2018/9೥ هࡌ͕૿߲͑ͨ໨ͷهࡌ͕͋Δ࿦จ਺ ਓ͔ؒΒֶͿɿ10%͘Β͍ ϢʔβʔΠϯλʔϑΣʔεɿ5%ҎԼ ࣗ཯ੑɺਖ਼ٛɺݸਓͷଚॏɿθϩ ͍ΘΏΔτοϓΧϯϑΝʹ͸͋·Γग़ͳ͍࿩୊

Slide 16

Slide 16 text

/ 86 ͓͜ͱΘΓ 02 u "* ʜ ػցֶशΛ͸͡Ίͱͨ͠਺ཧతͳΞϧΰϦζϜʹجͮ͘ख๏શൠ ࣗಈԽγεςϜͳͲͷ࿩୊΋ؚ·ΕΔ u ೝ஌όΠΞεʜߴ࣍ೝ஌ಛੑʹىҼͯ͠ɺਓؒͷҙࢥܾఆ͕ภΔ͜ͱ AI։ൃऀ͕ཧ૝తͩͱ૝ఆ͢ΔͰ͋Ζ͏ ҙࢥܾఆΛج४ͱͯ͠ ͜ͱ͹ͷఆٛ u "*͕ਓؒΛ׬શʹ୅ସ͢Δ৔໘ʢҙࢥܾఆʹਓ͕ؒҰ੾ؔ༩͠ͳ͍৔໘ʣ u ࠩผɺެฏੑɺσδλϧɾσΟόΠυ u ϑΣΠΫχϡʔεɺEJTJOGPSNBUJPO u ηΩϡϦςΟɺ৘ใ࿙Ӯɺݸਓ৘ใอޢ u ࣗવݴޠॲཧɾը૾ॲཧɾϩϘοτ΍ΤʔδΣϯτʹݶఆ͞Εͨ࿩ ࠓ೔͸࿩͞ͳ͍͜ͱ ஫ʣͱͯ΋޿͍ҙຯͰ࢖͍ͬͯ·͢ɻ

Slide 17

Slide 17 text

AI։ൃͷ্ྲྀաఔ

Slide 18

Slide 18 text

/ 86 Take Home Message 1 10 ਺஋ʹͨ͠ॠؒʹҰਓา͖ͯ͠͠·͏΋ͷ͸ੈʹଟ͋͘Δɻ ηΦυΞɾMɾϙʔλʔஶ ౻֞༟ࢠ༁ ʮ਺஋ͱ٬؍ੑʯ p.299

Slide 19

Slide 19 text

/ 86 ͸͡Ίʹ 10 ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ

Slide 20

Slide 20 text

/ 86 ͸͡Ίʹ 10 ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ train test ڭࢣ͋Γ σʔληοτ ໨࣍ σʔλ࡞੒ͱೝ஌όΠΞε Ø ݸਓͰͷҙࢥܾఆ Ø ूஂతҙࢥܾఆ Ø σʔλԽʹΑΔ৘ใͷܽམ ղܾͷͨΊͷݚڀ Ø ٕज़తͳհೖ Ø ೝ஌ಛੑΛ௚઀తʹѻ͏հೖ Ø ͦͷଞ u ܭଌͷෆ͔֬͞ɺܭଌٕज़։ൃ ᵋ #.*ٕज़ɺηϯαʔ։ൃɺͳͲ u σʔλɾ৘ใަ׵ن໿ͷඪ४Խ ࠓ೔͸࿩͞ͳ͍͜ͱ

Slide 21

Slide 21 text

/ 86 ໨࣍ 10 σʔλ࡞੒ͱೝ஌όΠΞε Ø ݸਓͰͷҙࢥܾఆ Ø ूஂతҙࢥܾఆ Ø σʔλԽʹΑΔ৘ใͷܽམ Ø ҉໧తͳ஌Λѻ͍͍ͨ ղܾͷͨΊͷݚڀ Ø ٕज़తͳհೖ Ø ೝ஌ಛੑΛ௚઀తʹѻ͏հೖ Ø ͦͷଞ

Slide 22

Slide 22 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ1 11 *;NE0@'-7C3?-( FJIDGL:38''9 :38>@/- %"&"! ;6 =1+A<40)+A<40) ؾ෼ͷམͪࠐΈ͕ ͋Γ·͔͢ʁ 0(ͳ͍)ʙ100(͋Δ) Ͱճ౴͍ͯͩ͘͠͞ ؾ෼ͷམͪࠐΈ͕ ͋Γ·͔͢ʁ 0(ͳ͍)ʙ100(͋Δ) Ͱճ౴͍ͯͩ͘͠͞ ؾ෼ͷམͪࠐΈ͕ ͋Γ·͔͢ʁ 0(ͳ͍)ʙ100(͋Δ) Ͱճ౴͍ͯͩ͘͠͞ 0 100 0 100 50

Slide 23

Slide 23 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ1 11 WJTVBMQSFTFOUBUJPOSPVOEOVNCFS74QSFDJTFOVNCFSͳͲ u ཭ࢄई౓Likertई౓͸Կஈ֊ʹ͢΂͖͔ [2] u ਺஋هࡌΩϦͷ͍͍਺ࣈ(round number) ʹภΔ [3] u બ୒ࢶઃܭʹΑΓճ౴͕࿪Ή[4] [1] [1] Matejka, J., Glueck, M., Grossman, T., & Fitzmaurice, G. (2016, May). The effect of visual appearance on the performance of continuous sliders and visual analogue scales. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (pp. 5421-5432). [2] Leung, S. O. (2011). A comparison of psychometric properties and normality in 4-, 5-, 6-, and 11-point Likert scales. Journal of social service research, 37(4), 412-421. [3] Jansen, C. J., & Pollmann, M. M. (2001). On round numbers: Pragmatic aspects of numerical expressions. Journal of quantitative linguistics, 8(3), 187-201. [4] ૿ాΒɺ৺ཧֶ͕ඳ͘ϦεΫͷੈք Advancedɺܚጯٛक़େֶग़൛ձɺ2023ɺp.12-21

Slide 24

Slide 24 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 #$72A5,0@ : KHMC 38/,8

Slide 25

Slide 25 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 https://www.toyo.co.jp/medica l/casestudy/detail/id=5525 https://github.com/ieee8023/c ovid-chestxray-dataset #$72A5,0@ : KHMC 38/,8 ※εϥΠυʹهࡌ ͨ͠ϥϕϧ͸ྫͰ͢

Slide 26

Slide 26 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 https://www.toyo.co.jp/medica l/casestudy/detail/id=5525 https://github.com/ieee8023/c ovid-chestxray-dataset ※εϥΠυʹهࡌ ͨ͠ϥϕϧ͸ྫͰ͢ #$72A5,0@ : KHMC 38/,8

Slide 27

Slide 27 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 https://www.toyo.co.jp/medica l/casestudy/detail/id=5525 https://github.com/ieee8023/c ovid-chestxray-dataset ※εϥΠυʹهࡌ ͨ͠ϥϕϧ͸ྫͰ͢ #$72A5,0@ : KHMC 38/,8

Slide 28

Slide 28 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 https://www.toyo.co.jp/medica l/casestudy/detail/id=5525 https://github.com/ieee8023/c ovid-chestxray-dataset ※εϥΠυʹهࡌ ͨ͠ϥϕϧ͸ྫͰ͢ B.5

Slide 29

Slide 29 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 https://www.toyo.co.jp/medica l/casestudy/detail/id=5525 https://github.com/ieee8023/c ovid-chestxray-dataset ※εϥΠυʹهࡌ ͨ͠ϥϕϧ͸ྫͰ͢ B.5

Slide 30

Slide 30 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 ϓϥΠϛϯάʢQSJNJOHʣΞϯΧϦϯάʢBODIPSJOHʣ [3] ਓؒͷ൓Ԡ͕ɺલͷܹࢗʹӨڹΛड͚Δ < > [1] Meyer, D. E., & Schvaneveldt, R. W. (1971). Facilitation in recognizing pairs of words: evidence of a dependence between retrieval operations. Journal of experimental psychology, 90(2), 227. [2] Tversky, A., & Kahneman, D., (1974). Judgement under uncertainty: Heuristics and biases. Science, 185, 1124-1131. [3] Koehler, D. J., & Harvey, N. (Eds.). (2008). Blackwell handbook of judgment and decision making. John Wiley & Sons., p.99 [3] Valdez, A. C., Ziefle, M., & Sedlmair, M. (2017). Priming and anchoring effects in visualization. IEEE transactions on visualization and computer graphics, 24(1), 584-594.

Slide 31

Slide 31 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 ϓϥΠϛϯάʢQSJNJOHʣΞϯΧϦϯάʢBODIPSJOHʣ [3] ਓؒͷ൓Ԡ͕ɺલͷܹࢗʹӨڹΛड͚Δ < > [1] Meyer, D. E., & Schvaneveldt, R. W. (1971). Facilitation in recognizing pairs of words: evidence of a dependence between retrieval operations. Journal of experimental psychology, 90(2), 227. [2] Tversky, A., & Kahneman, D., (1974). Judgement under uncertainty: Heuristics and biases. Science, 185, 1124-1131. [3] Koehler, D. J., & Harvey, N. (Eds.). (2008). Blackwell handbook of judgment and decision making. John Wiley & Sons., p.99 [3] Valdez, A. C., Ziefle, M., & Sedlmair, M. (2017). Priming and anchoring effects in visualization. IEEE transactions on visualization and computer graphics, 24(1), 584-594.

Slide 32

Slide 32 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ2 12 ϓϥΠϛϯάʢQSJNJOHʣΞϯΧϦϯάʢBODIPSJOHʣ ਓؒͷ൓Ԡ͕ɺલͷܹࢗʹӨڹΛड͚Δ < > [1] Meyer, D. E., & Schvaneveldt, R. W. (1971). Facilitation in recognizing pairs of words: evidence of a dependence between retrieval operations. Journal of experimental psychology, 90(2), 227. [2] Tversky, A., & Kahneman, D., (1974). Judgement under uncertainty: Heuristics and biases. Science, 185, 1124-1131. [3] Koehler, D. J., & Harvey, N. (Eds.). (2008). Blackwell handbook of judgment and decision making. John Wiley & Sons., p.99 [3] Valdez, A. C., Ziefle, M., & Sedlmair, M. (2017). Priming and anchoring effects in visualization. IEEE transactions on visualization and computer graphics, 24(1), 584-594. [4]

Slide 33

Slide 33 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ3 13 ·ͱΊ A B A A

Slide 34

Slide 34 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ3 13 ·ͱΊ A B A A A B ͋ͬ… Bͬ͢ϋΠ B A ͩͬ ͭͬͯΜͩΖΦϥ Aාͬ Aාͬ AશձҰக͢͝ʔ͍

Slide 35

Slide 35 text

/ 86 ࡞੒͞ΕΔσʔλͷภΓ3 13 ܦ࿏ґଘੑಉௐ ूஂ ͳͲ u ҙݟू໿ ◇ ܦ࿏ґଘੑʢpath dependenceʣʢΞϩʔͷෆՄೳੑఆཧʣ[1] ◇ ಉௐʢconformityʣ[2] ݸਓͷ൑அ͕ಉ੮͢Δଞऀͷ൑அʹӨڹΛड͚Δ ◇ ूஂۃੑԽʢ੒ۃԽʣʢgroup polarizationʣ[3] ूஂʹΑΔٞ࿦‎ҙݟ͕ۃ୺ʢաܹ/৻ॏʣʹͳΔ [1] Arrow, Kenneth J. (1950). "A Difficulty in the Concept of Social Welfare" . Journal of Political Economy. 58 (4): 328–346. [2] Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgments. [3] Moscovici, S., & Zavalloni, M. (1969). The group as a polarizer of attitudes. Journal of personality and social psychology, 12(2), 125. [4] ޿ాΒɺ৺ཧֶ͕ඳ͘ϦεΫͷੈքɺܚጯٛक़େֶग़൛ձɺp.179 [5] ന઒ల೭, & খࣲ. (2018). Պֶٕज़༧ଌௐࠪख๏ʹؔ͢Δ਺ཧత෼ੳ σϧϑΝΠௐࠪٴͼϦΞϧλΠϜɾσϧϑΝΠ๏ʹؔ͢ΔΤʔδΣϯτγϛϡϨʔγϣϯ. ݚڀ ٕज़ ܭը, 33(2), 170-183. u σϧϑΝΠؔ࿈ख๏ ख๏ʹґଘͯ͠ɺ࠷ऴతʹू໿͞ΕΔҙݟ͕ҟͳΔ [5] ҙݟू໿ͷ ܾఆϧʔϧ ܾఆ݁Ռʹ͍ͭͯͷ֤Ϟσϧͷ༧ଌ஋ ༗ࡑ ແࡑ ະܾ ൺྫܕ .220 .780 .000 2/3ଟ਺ܾܕ .001 .899 .100 ຬ৔Ұகܕ .000 1.000 .000 [4]

Slide 36

Slide 36 text

/ 86 σʔλԽʹΑΔ৘ใͷܽམ1 ա౓ͳ୯७Խ 14 & *$#' %(")! A@H?5D6J> 9711J>9,;2; 2 5>@ BA2=B C@AE 8@.+A6/

Slide 37

Slide 37 text

/ 86 σʔλԽʹΑΔ৘ใͷܽམ1 ա౓ͳ୯७Խ 14 Suresh, H., & Guttag, J. (2021). A framework for understanding sources of harm throughout the machine learning life cycle. In Equity and access in algorithms, mechanisms, and optimization (pp. 1-9). & *$#' %(")! lDCJd A@H?5D6J> 9711J>9,;2; 2 5>@ BA2=B C@AE 8@.+A6/

Slide 38

Slide 38 text

/ 86 σʔλԽʹΑΔ৘ใͷܽམ2 Ұҙʹఆ·Βͳ͍஌ 15 Krauss, J. C., Boonstra, P. S., Vantsevich, A. V., & Friedman, C. P. (2016). Is the problem list in the eye of the beholder? An exploration of consistency across physicians. Journal of the American Medical Informatics Association, 23(5), 859-865. Kagawa, R., Shinohara, E., Imai, T., Kawazoe, Y., & Ohe, K. (2019). Bias of inaccurate disease mentions in electronic health record-based phenotyping. International Journal of Medical Informatics, 124, 90-96. A@H?5D6J> 9711J>9,;2; 2 5>@ BA2=B C@AE 8@.+A6/ & *$#' %(")! lDCJd Y]dDC; f9 .ÆÜÃÙ";

Slide 39

Slide 39 text

/ 86 ղܾʹ޲͚ͨࢼΈ 16 ٕ ज़ ։ ൃ u φοδҙࢥܾఆΛಛఆͷํ޲ʹม͑Δબ୒ΞʔΩςΫνϟͷཁૉ <> u ϒʔετೝ஌తٕྔΛߴΊΔ·ͨ͸ ֫ಘ͢Δ͜ͱͰ߹ཧతͳҙࢥܾఆʹಋ͘<> u ख़ߟΛଅ͢ <> ਓ ؒ ʹ հ ೖ AI UI u ೝ஌όΠΞεͷӨڹΛܰݮͨ͠ϥϕϧΛ༧ଌ͢Δ਺ཧϞσϧ [3, 4] u ೝ஌όΠΞεܰݮͷͨΊͷΫϥ΢υιʔγϯάͰͷνΣοΫϦετ [11] u ϊΠζͷ͋Δσʔλʹରͯ͠ؤ݈ͳػցֶशϞσϧ [5] u গ਺σʔλʹରͯ͠ؤ݈ͳػցֶशϞσϧ few-shot learning, semi/weakly-supervised, domain adaptation, self-supervised, … u Human-in-the-Loop ػցֶश [6] u VASͷઃܭͷఏҊ [1] ໨੝Γºɺ஋͕ݟ͑ΔಈతεϥΠμʔ ̋ɺ ଳঢ়ϥϕϧ2ͭͷεϥΠμʔ ̋ u ௥Ճઃ໰΍৘ใఏࣔͰରॲ [2] [1] Matejka, J., Glueck, M., Grossman, T., & Fitzmaurice, G. (2016, May). The effect of visual appearance on the performance of continuous sliders and visual analogue scales. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (pp. 5421-5432). [2] Hube, C., Fetahu, B., & Gadiraju, U. (2019, May). Understanding and mitigating worker biases in the crowdsourced collection of subjective judgments. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (pp. 1-12). [3] Zhuang, H., Parameswaran, A., Roth, D., & Han, J. (2015, August). Debiasing crowdsourced batches. In Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 1593-1602). [4] Gemalmaz, M. A., & Yin, M. (2021). Accounting for Confirmation Bias in Crowdsourced Label Aggregation. In IJCAI (pp. 1729-1735). [5] Song, H., Kim, M., Park, D., Shin, Y., & Lee, J. G. (2022). Learning from noisy labels with deep neural networks: A survey. IEEE Transactions on Neural Networks and Learning Systems [6] .Mosqueira-Rey, E., Hernández-Pereira, E., Alonso-Ríos, D., Bobes-Bascarán, J., & Fernández-Leal, Á. (2023). Human-in-the-loop machine learning: A state of the art. Artificial Intelligence Review, 56(4), 3005-3054. .[11] Draws, T., Rieger, A., Inel, O., Gadiraju, U., & Tintarev, N. (2021, October). A checklist to combat cognitive biases in crowdsourcing. In Proceedings of the AAAI conference on human computation and crowdsourcing (Vol. 9, pp. 48-59). [7] Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth and happiness. Simon & Schuster [8] Hertwig, R., & Grüne-Yanoff, T. (2017). Nudging and boosting: Steering or empowering good decisions. Perspectives on Psychological Science, 12 (6), 973–986. [9] O’Sullivan, E. D., & Schofield, S. J. (2019). A cognitive forcing tool to mitigate cognitive bias–a randomised control trial. BMC medical education, 19, 1-8.[10] Kameda, T., Toyokawa, W., & Tindale, R. S. (2022). Information aggregation and collective intelligence beyond the wisdom of crowds. Nature Reviews Psychology, 1(6), 345-357 ू ஂ ݸ ਓ u ूஂతҙࢥܾఆʹΑΓਖ਼֬ੑ͕޲্͢Δ৚݅ͷ໢ཏతݕ౼ <>

Slide 40

Slide 40 text

/ 86 ղܾʹ޲͚ͨࢼΈ 16 ٕ ज़ ։ ൃ u φοδҙࢥܾఆΛಛఆͷํ޲ʹม͑Δબ୒ΞʔΩςΫνϟͷཁૉ <> u ϒʔετೝ஌తٕྔΛߴΊΔ·ͨ͸ ֫ಘ͢Δ͜ͱͰ߹ཧతͳҙࢥܾఆʹಋ͘<> u ख़ߟΛଅ͢ <> ਓ ؒ ʹ հ ೖ AI UI u ೝ஌όΠΞεͷӨڹΛܰݮͨ͠ϥϕϧΛ༧ଌ͢Δ਺ཧϞσϧ [3, 4] u ೝ஌όΠΞεܰݮͷͨΊͷΫϥ΢υιʔγϯάͰͷνΣοΫϦετ [11] u ϊΠζͷ͋Δσʔλʹରͯ͠ؤ݈ͳػցֶशϞσϧ [5] u গ਺σʔλʹରͯ͠ؤ݈ͳػցֶशϞσϧ few-shot learning, semi/weakly-supervised, domain adaptation, self-supervised, … u Human-in-the-Loop ػցֶश [6] u VASͷઃܭͷఏҊ [1] ໨੝Γºɺ஋͕ݟ͑ΔಈతεϥΠμʔ ̋ɺ ଳঢ়ϥϕϧ2ͭͷεϥΠμʔ ̋ u ௥Ճઃ໰΍৘ใఏࣔͰରॲ [2] [1] Matejka, J., Glueck, M., Grossman, T., & Fitzmaurice, G. (2016, May). The effect of visual appearance on the performance of continuous sliders and visual analogue scales. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (pp. 5421-5432). [2] Hube, C., Fetahu, B., & Gadiraju, U. (2019, May). Understanding and mitigating worker biases in the crowdsourced collection of subjective judgments. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (pp. 1-12). [3] Zhuang, H., Parameswaran, A., Roth, D., & Han, J. (2015, August). Debiasing crowdsourced batches. In Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 1593-1602). [4] Gemalmaz, M. A., & Yin, M. (2021). Accounting for Confirmation Bias in Crowdsourced Label Aggregation. In IJCAI (pp. 1729-1735). [5] Song, H., Kim, M., Park, D., Shin, Y., & Lee, J. G. (2022). Learning from noisy labels with deep neural networks: A survey. IEEE Transactions on Neural Networks and Learning Systems [6] .Mosqueira-Rey, E., Hernández-Pereira, E., Alonso-Ríos, D., Bobes-Bascarán, J., & Fernández-Leal, Á. (2023). Human-in-the-loop machine learning: A state of the art. Artificial Intelligence Review, 56(4), 3005-3054. .[11] Draws, T., Rieger, A., Inel, O., Gadiraju, U., & Tintarev, N. (2021, October). A checklist to combat cognitive biases in crowdsourcing. In Proceedings of the AAAI conference on human computation and crowdsourcing (Vol. 9, pp. 48-59). [7] Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth and happiness. Simon & Schuster [8] Hertwig, R., & Grüne-Yanoff, T. (2017). Nudging and boosting: Steering or empowering good decisions. Perspectives on Psychological Science, 12 (6), 973–986. [9] O’Sullivan, E. D., & Schofield, S. J. (2019). A cognitive forcing tool to mitigate cognitive bias–a randomised control trial. BMC medical education, 19, 1-8.[10] Kameda, T., Toyokawa, W., & Tindale, R. S. (2022). Information aggregation and collective intelligence beyond the wisdom of crowds. Nature Reviews Psychology, 1(6), 345-357 ू ஂ ݸ ਓ u ूஂతҙࢥܾఆʹΑΓਖ਼֬ੑ͕޲্͢Δ৚݅ͷ໢ཏతݕ౼ <>

Slide 41

Slide 41 text

/ 86 ղܾʹ޲͚ͨࢼΈ 16 ٕ ज़ ։ ൃ u φοδҙࢥܾఆΛಛఆͷํ޲ʹม͑Δબ୒ΞʔΩςΫνϟͷཁૉ <> u ϒʔετೝ஌తٕྔΛߴΊΔ·ͨ͸ ֫ಘ͢Δ͜ͱͰ߹ཧతͳҙࢥܾఆʹಋ͘<> u ख़ߟΛଅ͢ <> ਓ ؒ ʹ հ ೖ AI UI u ೝ஌όΠΞεͷӨڹΛܰݮͨ͠ϥϕϧΛ༧ଌ͢Δ਺ཧϞσϧ [3, 4] u ೝ஌όΠΞεܰݮͷͨΊͷΫϥ΢υιʔγϯάͰͷνΣοΫϦετ [11] u ϊΠζͷ͋Δσʔλʹରͯ͠ؤ݈ͳػցֶशϞσϧ [5] u গ਺σʔλʹରͯ͠ؤ݈ͳػցֶशϞσϧ few-shot learning, semi/weakly-supervised, domain adaptation, self-supervised, … u Human-in-the-Loop ػցֶश [6] u VASͷઃܭͷఏҊ [1] ໨੝Γºɺ஋͕ݟ͑ΔಈతεϥΠμʔ ̋ɺ ଳঢ়ϥϕϧ2ͭͷεϥΠμʔ ̋ u ௥Ճઃ໰΍৘ใఏࣔͰରॲ [2] [1] Matejka, J., Glueck, M., Grossman, T., & Fitzmaurice, G. (2016, May). The effect of visual appearance on the performance of continuous sliders and visual analogue scales. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (pp. 5421-5432). [2] Hube, C., Fetahu, B., & Gadiraju, U. (2019, May). Understanding and mitigating worker biases in the crowdsourced collection of subjective judgments. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (pp. 1-12). [3] Zhuang, H., Parameswaran, A., Roth, D., & Han, J. (2015, August). Debiasing crowdsourced batches. In Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 1593-1602). [4] Gemalmaz, M. A., & Yin, M. (2021). Accounting for Confirmation Bias in Crowdsourced Label Aggregation. In IJCAI (pp. 1729-1735). [5] Song, H., Kim, M., Park, D., Shin, Y., & Lee, J. G. (2022). Learning from noisy labels with deep neural networks: A survey. IEEE Transactions on Neural Networks and Learning Systems [6] .Mosqueira-Rey, E., Hernández-Pereira, E., Alonso-Ríos, D., Bobes-Bascarán, J., & Fernández-Leal, Á. (2023). Human-in-the-loop machine learning: A state of the art. Artificial Intelligence Review, 56(4), 3005-3054. .[11] Draws, T., Rieger, A., Inel, O., Gadiraju, U., & Tintarev, N. (2021, October). A checklist to combat cognitive biases in crowdsourcing. In Proceedings of the AAAI conference on human computation and crowdsourcing (Vol. 9, pp. 48-59). [7] Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving decisions about health, wealth and happiness. Simon & Schuster [8] Hertwig, R., & Grüne-Yanoff, T. (2017). Nudging and boosting: Steering or empowering good decisions. Perspectives on Psychological Science, 12 (6), 973–986. [9] O’Sullivan, E. D., & Schofield, S. J. (2019). A cognitive forcing tool to mitigate cognitive bias–a randomised control trial. BMC medical education, 19, 1-8.[10] Kameda, T., Toyokawa, W., & Tindale, R. S. (2022). Information aggregation and collective intelligence beyond the wisdom of crowds. Nature Reviews Psychology, 1(6), 345-357 ू ஂ ݸ ਓ u ूஂతҙࢥܾఆʹΑΓਖ਼֬ੑ͕޲্͢Δ৚݅ͷ໢ཏతݕ౼ <>

Slide 42

Slide 42 text

/ 86 ·ͱΊ 19 Θ͔͍ͬͯΔ͜ͱɾͰ͖͍ͯΔ͜ͱ ͜Ε͔Βͷ՝୊ u AIɾσʔλαΠΤϯε෼໺ʹ͓͚Δೝ஌όΠΞεͷཧղ ◇ ఆྔతͳٞ࿦ ◇ ଟ෼໺ʹ·͕ͨΔ஌ݟͷ౷߹తཧղ u ೝ஌όΠΞεͷܰݮʹ޲͚ͨԠ༻ݚڀ ◇ AIͷΞϧΰϦζϜͷఏҊ ◇ UIͷఏҊ ◇ ҙࢥܾఆաఔʹհೖ͢Δํ๏࿦ͷఏҊ u ࡞ۀ؀ڥʹىҼͯ͠ɺ࡞੒͞ΕͨσʔλʹόΠΞε͕ੜ͡͏Δ u ࡞ۀ؀ڥͱ͸ؔ܎ͳͯ͘΋ɺ࡞੒͞ΕͨσʔλʹόΠΞε͕ੜ͡͏Δ _S ¸As ²MS#¿ I¥½´3L»¼¾¡­³°™

Slide 43

Slide 43 text

AI։ൃͷԼྲྀաఔ

Slide 44

Slide 44 text

/ 86 Take Home Message 2 20 ͋͋ݴ͑͹ɺ͜͏ݴ͏ɻ

Slide 45

Slide 45 text

/ 86 ͸͡Ίʹ 20 ਆͷΈͧ஌Δ ਅͷੈք σʔλੜ੒ α ϯ ϓ Ϧ ϯ ά ͜ͷ࢟΋ ਆͷΈͧ஌Δ ܭଌ ٬ ମ Խ σʔληοτ લ ॲ ཧ ࣗಈతॲཧ ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ train test ڭࢣ͋Γ σʔληοτ ※ લఏɿAIͷੑೳ͸ۃΊͯߴ͍

Slide 46

Slide 46 text

/ 86 ͸͡Ίʹ 20 ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ ※ લఏɿAIͷੑೳ͸ۃΊͯߴ͍ u ҩྍ਍அࢧԉ"* ྟচݚڀͰɺҩऀͷύϑΥʔϚϯε΍ױऀͷసػ͕վળͨ͠ใࠂ͕΄΅ͳ͍ɻ< > [1] Vasey, B., Ursprung, S., Beddoe, B., Taylor, E. H., Marlow, N., Bilbro, N., ... & McCulloch, P. (2021). Association of clinician diagnostic performance with machine learning–based decision support systems: a systematic review. JAMA network open, 4(3), e211276-e211276. [2] Freeman, K., Geppert, J., Stinton, C., Todkill, D., Johnson, S., Clarke, A., & Taylor-Phillips, S. (2021). Use of artificial intelligence for image analysis in breast cancer screening programmes: systematic review of test accuracy. bmj, 374.

Slide 47

Slide 47 text

/ 86 ͸͡Ίʹ 20 ֶशɾਪ࿦ ޙ ॲ ཧ γ ε ς Ϝ Խ ධՁ Ϟσϧ train test ڭࢣ͋Γ σʔληοτ [1] Lai, V., Chen, C., Smith-Renner, A., Liao, Q. V., & Tan, C. (2023, June). Towards a Science of Human-AI Decision Making: An Overview of Design Space in Empirical Human-Subject Studies. In Pr of the 2023 ACM Conference on Fairness, Accountability, and Transparency (pp. 1369-1385). [1] ಓಙత൑அ΍૑଄తͳҙࢥܾఆ ʢJF ౴͕͑ఆ·Βͳ͍λεΫʣ ࠓ೔͸࿩͞ͳ͍͜ͱ ※ લఏɿAIͷੑೳ͸ۃΊͯߴ͍ "*͕༧ଌ·ͨ͸ࣝผͨ݁͠ՌΛɺਓؒ͸ͲͷΑ͏ʹ࢖ͬͯҙࢥܾఆ͢Δͷ͔ʁ

Slide 48

Slide 48 text

/ 86 ໨࣍ 20 ਓؒ͸"*Λద੾ʹ࢖͑ͳ͍ Ø "*ͷաখධՁͱաେධՁ ղܾͷج൫ͱͳΓ͏Δݚڀ Ø "*ٕज़։ൃܥ Ø આ໌ՄೳੑɾղऍՄೳੑ Ø ৴པɺઆಘ Ø ػցֶशΞϧΰϦζϜͷ։ൃ Ø ਓؒͷҙࢥܾఆաఔΛཧղ͢Δݚڀ

Slide 49

Slide 49 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 աখධՁ "*Λॿݴͱͯ͠ར༻ ࣾձ৺ཧֶ +VEHF"EWJTPS4ZTUFN +"4 ʹै͏ o ҩྍ਍அࢧԉ o ࠶൜༧ଌ ͳͲ "*Λ؂ࢹ͢Δ ΤϧΰϊϛΫεɾώϡʔϚϯϑΝΫλʔ o ඈߦػͷࣗಈૢॎ o ंͷࣗಈӡస Ϩϕϧ2ʙ4 o ޻৔಺Ͱͷނোݕ஌ ͳͲ ࣗݾத৺త ॿݴׂҾ <> ࠷ॳʹਓ͔"*͔ΛબͿ ϚʔέςΟϯά o ঎඼ߪೖɾ޿ࠂ ͳͲ ࣗಈԽγεςϜ EJTVTF<> NJTVTF<> BMHPSJUINBWFSTJPO ớ Ξ ϧ ΰ Ϧ ζ Ϝ ݏ ѱ Ờ < > DPNQMBDFODZ <> աେධՁ NBDIJOFIFVSJTUJDT< > QFSGFDUBVUPNBUJPOTDIFNB<> BVUPNBUJPO CJBT<> advice taking [1] Yaniv, I., & Kleinberger, E. (2000). Advice taking in decision making: Egocentric discounting and reputation formation. Organizational behavior and human decision processes, 83(2), 260-281 [2] Dietvorst, B. J., Simmons, J. P., & Massey, C. (2015). Algorithm aversion: people erroneously avoid algorithms after seeing them err. Journal of experimental psychology: General, 144(1), 114. [3] S. Shyam Sundar and Jinyoung Kim. 2019. Machine Heuristic: When We Trust Computers More than Humans with Our Personal Information. CHI2019, p.1-9 [4] Parasuraman, R., & Riley, V. (1997). Humans and automation: Use, misuse, disuse, abuse. Human factors, 39(2), 230-253. [5] Parasuraman, R., & Manzey, D. H. (2010). Complacency and bias in human use of automation: An attentional integration. Human factors, 52(3), 381-410. [6] Skitka, L. J., Mosier, K. L., & Burdick, M. (1999). Does automation bias decision-making?. International Journal of Human-Computer Studies, 51(5), 991-1006. [7] Tschopp, M. (2020). PAS-The perfect automation schema. https://www.scip.ch/en/?labs.20200507

Slide 50

Slide 50 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 Chen, C., Feng, S., Sharma, A., & Tan, C. (2022). Machine Explanations and Human Understanding. Transactions on Machine Learning Research. ਅͷڥք ਓ͕΋ͱ΋ͱ ͍࣋ͬͯͨڥք ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ AIͷڥք ʔ ʔ ʴ

Slide 51

Slide 51 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 Chen, C., Feng, S., Sharma, A., & Tan, C. (2022). Machine Explanations and Human Understanding. Transactions on Machine Learning Research. ਅͷڥք ਓ͕΋ͱ΋ͱ ͍࣋ͬͯͨڥք ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ AIͷڥք ʔ ʔ ʴ

Slide 52

Slide 52 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 Chen, C., Feng, S., Sharma, A., & Tan, C. (2022). Machine Explanations and Human Understanding. Transactions on Machine Learning Research. ਅͷڥք ਓ͕΋ͱ΋ͱ ͍࣋ͬͯͨڥք ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ AIͷڥք ʔ ʔ ʴ

Slide 53

Slide 53 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 Chen, C., Feng, S., Sharma, A., & Tan, C. (2022). Machine Explanations and Human Understanding. Transactions on Machine Learning Research. ਅͷڥք ਓ͕΋ͱ΋ͱ ͍࣋ͬͯͨڥք ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ AIͷڥք ʔ ʔ ʴ ʔ

Slide 54

Slide 54 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 Chen, C., Feng, S., Sharma, A., & Tan, C. (2022). Machine Explanations and Human Understanding. Transactions on Machine Learning Research. ਅͷڥք ਓ͕΋ͱ΋ͱ ͍࣋ͬͯͨڥք ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ AIͷڥք ʔ ʔ ʴ ͜ͷ͋ͨΓ͸AIͷݴ͏௨Γʹͨ͠ํ͕͍͍ʢAIͷաখධՁ͕໰୊ʣ

Slide 55

Slide 55 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 Chen, C., Feng, S., Sharma, A., & Tan, C. (2022). Machine Explanations and Human Understanding. Transactions on Machine Learning Research. ਅͷڥք ਓ͕΋ͱ΋ͱ ͍࣋ͬͯͨڥք ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ AIͷڥք ʔ ʔ ʴ ͜ͷ͋ͨΓ͸ࣗ෼ͷߟ͑Λ৴ͨ͡ํ͕͍͍ʢAIͷաେධՁ͕໰୊ʣ ʔ

Slide 56

Slide 56 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 Chen, C., Feng, S., Sharma, A., & Tan, C. (2022). Machine Explanations and Human Understanding. Transactions on Machine Learning Research. ਅͷڥք ਓ͕΋ͱ΋ͱ ͍࣋ͬͯͨڥք ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ AIͷڥք ʔ ʔ ʴ ʔ ্3ͭͷઢͷ͢΂ͯΛਓ͕ؒ ೺ѲͰ͖ͳ͍ͱ ਓؒ͸AIΛద੾ʹ࢖͑ͳ͍

Slide 57

Slide 57 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 աখධՁ "*Λॿݴͱͯ͠ར༻ ࣾձ৺ཧֶ +VEHF"EWJTPS4ZTUFN +"4 ʹै͏ o ҩྍ਍அࢧԉ o ࠶൜༧ଌ ͳͲ "*Λ؂ࢹ͢Δ ΤϧΰϊϛΫεɾώϡʔϚϯϑΝΫλʔ o ඈߦػͷࣗಈૢॎ o ंͷࣗಈӡస Ϩϕϧ2ʙ4 o ޻৔಺Ͱͷނোݕ஌ ͳͲ ࣗݾத৺త ॿݴׂҾ <> ࠷ॳʹਓ͔"*͔ΛબͿ ϚʔέςΟϯά o ঎඼ߪೖɾ޿ࠂ ͳͲ ࣗಈԽγεςϜ EJTVTF<> NJTVTF<> BMHPSJUINBWFSTJPO ớ Ξ ϧ ΰ Ϧ ζ Ϝ ݏ ѱ Ờ < > DPNQMBDFODZ <> աେධՁ NBDIJOFIFVSJTUJDT< > QFSGFDUBVUPNBUJPOTDIFNB<> BVUPNBUJPO CJBT<> advice taking [1] Yaniv, I., & Kleinberger, E. (2000). Advice taking in decision making: Egocentric discounting and reputation formation. Organizational behavior and human decision processes, 83(2), 260-281 [2] Dietvorst, B. J., Simmons, J. P., & Massey, C. (2015). Algorithm aversion: people erroneously avoid algorithms after seeing them err. Journal of experimental psychology: General, 144(1), 114. [3] S. Shyam Sundar and Jinyoung Kim. 2019. Machine Heuristic: When We Trust Computers More than Humans with Our Personal Information. CHI2019, p.1-9 [4] Parasuraman, R., & Riley, V. (1997). Humans and automation: Use, misuse, disuse, abuse. Human factors, 39(2), 230-253. [5] Parasuraman, R., & Manzey, D. H. (2010). Complacency and bias in human use of automation: An attentional integration. Human factors, 52(3), 381-410. [6] Skitka, L. J., Mosier, K. L., & Burdick, M. (1999). Does automation bias decision-making?. International Journal of Human-Computer Studies, 51(5), 991-1006. [7] Tschopp, M. (2020). PAS-The perfect automation schema. https://www.scip.ch/en/?labs.20200507

Slide 58

Slide 58 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 աখධՁ ΞϧΰϦζϜݏѱʢBMHPSJUINBWFSTJPOʣ "*΋ؒҧ͑Δ͜ͱΛ஌Δͱɺͦͷ"*ͷग़ྗΑΓ΋ࣗ෼ͷ൑அΛબ୒͢Δ <> ୠɿʮਓ͕ؒࣗ෼ΑΓ΋ਫ਼౓ͷߴ͍ΞϧΰϦζϜͷग़ྗΛ࢖Θͳ͍ʯ ͷҙຯͰ࢖ΘΕΔ͜ͱ͕ଟ͍ 0 100 200 300 400 500 600 700 800 900 1000 2015 2016 2017 2018 2019 2020 2021 2022 2023 GOOGLE SCHOLAR هࣄ਺ 2023/11/11ݱࡏ ↔ algorithm appreciation ਓؒ͸ΞϧΰϦζϜͷग़ྗΛ࠾༻͢Δ [2] [1] Dietvorst, B. J., Simmons, J. P., & Massey, C. (2015). Algorithm aversion: people erroneously avoid algorithms after seeing them err. Journal of experimental psychology: General, 144(1), 114. [2] Logg, J. M., Minson, J. A., & Moore, D. A. (2019). Algorithm appreciation: People prefer algorithmic to human judgment. Organizational Behavior and Human Decision Processes, 151, 90-103. ࣮ݧϓϩτίϧ [1]

Slide 59

Slide 59 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 աখධՁ ࣗݾத৺తॿݴׂҾʢFHPDFOUSJDBEWJDFEJTDPVOUJOHʣ ࣗઆʹ౎߹Α͘ɺଞऀ͔ΒͷॿݴΛׂΓҾ͘ <> ΠΪϦε͕ࠃͱͯ͠ਪ঑͍ͯ͠Δ ਍அࢧԉAIͷॿݴΛҩࢣ͕աখධՁ [2] ౶೘පͷൃ঱֬཰ ※ਅͷਖ਼ղ:70% ྫ ਍ ࡯ "*ͷࢧԉ ˞༧ଌɾࣝผਫ਼౓͕ۃΊͯߴ͍લఏ ࠷ऴతͳҙࢥܾఆ ױऀ9 ҩऀ" 50% 70% ༧ଌޡࠩ 35% ҩऀ" 70% [1] Yaniv, I., & Kleinberger, E. (2000). Advice taking in decision making: Egocentric discounting and reputation formation. Organizational behavior and human decision processes, 83(2), 260-281. [2] Pálfi, B., Arora, K., & Kostopoulou, O. (2022). Algorithm-based advice taking and clinical judgement: impact of advice distance and algorithm information. Cognitive research: principles and implications, 7(1), 70. [3] Bailey, P. E., Leon, T., Ebner, N. C., Moustafa, A. A., & Weidemann, G. (2022). A meta-analysis of the weight of advice in decision-making. Current Psychology, 1-26. ॿݴऀʹΑΒͣ ؤ݈ʹ࠶ݱ [3] ˞͜ͷҰ࿈ͷաఔΛ+VEHFBEWJTPSTZTUFNʢ+"4ʣͱݺͿɻඪ४తͳ࣮ݧखॱͰ΋͋Δ

Slide 60

Slide 60 text

/ 86 AI΁ͷաখධՁɾաେධՁ 21 աେධՁ ࣗಈԽόΠΞεʢBVUPNBUJPOCJBTʣ ࣗಈԽγεςϜʹա౓ʹґଘ͢Δ܏޲ <> [1] Skitka, L. J., Mosier, K. L., & Burdick, M. (1999). Does automation bias decision-making?. International Journal of Human-Computer Studies, 51(5), 991-1006. [2] Alon-Barkat, S., & Busuioc, M. (2023). Human–AI interactions in public sector decision making:“automation bias” and “selective adherence” to algorithmic advice. Journal of Public Administration Research and Theory, 33(1), 153-169. [3] Kupfer, C., Prassl, R., Fleiß, J., Malin, C., Thalmann, S., & Kubicek, B. (2023). Check the box! How to deal with automation bias in AI-based personnel selection. Frontiers in Psychology, 14, 1118723. [4] Lyell, D., & Coiera, E. (2017). Automation bias and verification complexity: a systematic review. Journal of the American Medical Informatics Association, 24(2), 423-431. "*ͷจ຺ߦ੓ [2]ɺاۀਓࣄ [3]ɺҩྍ [4] ͳͲ ;ͩΜ͔ΒࢧԉγεςϜΛར༻ͯ͠൑அ͍ͯ͠ΔͱɺࢧԉγεςϜ͕ͳ͘ͳͬͨͱ͖ʹɺ ൑அʹඞཁͳঢ়گͷݟམͱ͕͠૿͑ͨɻ λεΫը໘ ܭث1 ܭث2 ࢧԉγεςϜ͕ Πϕϯτʹج͖ͮ ߦಈΛਪન ࢀՃऀͷ൒෼͸͕ۭ͜͜ཝ ࣮ݧϓϩτίϧ [1]

Slide 61

Slide 61 text

/ 86 ໨࣍ 20 1. ਓؒ͸AIΛద੾ʹ࢖͑ͳ͍ Ø AIͷաখධՁͱաେධՁ ղܾͷج൫ͱͳΓ͏Δݚڀ Ø "*ٕज़։ൃܥ Ø આ໌ՄೳੑɾղऍՄೳੑ Ø ৴པɺઆಘ Ø ػցֶशΞϧΰϦζϜͷ։ൃ Ø ਓؒͷҙࢥܾఆաఔΛཧղ͢Δݚڀ

Slide 62

Slide 62 text

/ 86 eXplainable AI ʢXAIɿઆ໌ՄೳAIʣ 22 എܠ AIͷ࣮Ԡ༻͕೉͍͠ҰҼʮAI͸ϒϥοΫϘοΫεͰ݁Ռɾཧ༝Λઆ໌ࠔ೉ʯ[1, 2] ར఺ AIͷग़ྗʢ༧ଌɾࣝผʣΛਓ͕ؒཧղͰ͖Δ͜ͱΛ໨తͱٕͨ͠ज़ ՝୊ આ໌ੑ [3] ɾղऍੑ [4] ɾಁ໌ੑ͸ɺඞͣ͠΋ਓؒͷҙࢥܾఆΛม͑ͳ͍ [5] ਓ͕ؒߦಈʹࢸΔաఔʹ͓͍ͯɺཧղͱҙࢥܾఆ͸ผ [1] JST ,ʢઓུϓϩϙʔβϧʣAIԠ༻γεςϜͷ҆શੑɾ৴པੑΛ֬อ͢Δ৽ੈ୅ιϑτ΢ΣΞ޻ֶͷཱ֬, https://www.jst.go.jp/crds/pdf/2018/SP/CRDS-FY2018-SP-03.pdf, 2018 [2] Molnar, C. (2018). A guide for making black box models explainable. URL: https://christophm. github. io/interpretable-ml-book, 2(3) [3] https://www.darpa.mil/program/explainable-artificial-intelligence [4] Ali, S., Abuhmed, T., El-Sappagh, S., Muhammad, K., Alonso-Moral, J. M., Confalonieri, R., ... & Herrera, F. (2023). Explainable Artificial Intelligence (XAI): What we know and what is left to attain Trustworthy Artificial Intelligence. Information Fusion, 99, 101805. [5] Weller, A. (2019). Transparency: motivations and challenges. In Explainable AI: interpreting, explaining and visualizing deep learning (pp. 23-40). Cham: Springer International Publishing. [6] https://github.com/shap/shap#citations [7] https://github.com/marcotcr/lime [8] Ribeiro, M. T., Singh, S., & Guestrin, C. (2016, August). " Why should i trust you?" Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining (pp. 1135-1144). આ໌Մೳੑ ·ͨ͸ ղऍՄೳੑ [6] [8] [7]

Slide 63

Slide 63 text

/ 86 eXplainable AI ʢXAIɿઆ໌ՄೳAIʣ 22 എܠ AIͷ࣮Ԡ༻͕೉͍͠ҰҼʮAI͸ϒϥοΫϘοΫεͰ݁Ռɾཧ༝Λઆ໌ࠔ೉ʯ[1, 2] ར఺ AIͷग़ྗʢ༧ଌɾࣝผʣΛਓ͕ؒཧղͰ͖Δ͜ͱΛ໨తͱٕͨ͠ज़ ՝୊ આ໌ੑɾղऍੑɾಁ໌ੑ͸ɺඞͣ͠΋ਓؒͷҙࢥܾఆΛʢ։ൃऀͷ૝ఆ௨Γʹʣ ม͑ͳ͍ [5] ਓ͕ؒߦಈʹࢸΔաఔʹ͓͍ͯɺཧղͱҙࢥܾఆ͸ผ આ໌Մೳੑ ·ͨ͸ ղऍՄೳੑ [10] https://asd.gsfc.nasa.gov/conferences/ai/program/003-XAIforNASA.pdf [20] Sørensen, K., Van den Broucke, S., Fullam, J., Doyle, G., Pelikan, J., Slonska, Z., ... & (HLS-EU) Consortium Health Literacy Project European. (2012). Health literacy and public health: a systematic review and integration of definitions and models. BMC public health, 12, 1-13. આ໌ʢཧղʣͱ "*ͷద੾ͳར༻͸ ඞͣ͠΋Ұக͠ͳ͍ [10] [20] ਓ ؒ ͕ ߦ ಈ ʹ ࢸ Δ ա ఔ ஌֮ ཧղ ධՁ ൑அɾҙࢥܾఆ ߦಈ

Slide 64

Slide 64 text

/ 86 eXplainable AI ʢXAIɿઆ໌ՄೳAIʣ 22 Chen, C., Feng, S., Sharma, A., & Tan, C. (2022). Machine Explanations and Human Understanding. Transactions on Machine Learning Research. ਅͷڥք ਓ͕΋ͱ΋ͱ ͍࣋ͬͯͨڥք ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʴ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ ʔ AIͷڥք ʔ ʔ ʴ ্3ͭͷઢͷ͢΂ͯΛਓ͕ؒ೺ ѲͰ͖ͳ͍ͱ ਓؒ͸AIΛద੾ʹ࢖͑ͳ͍ XAI͕ղܾͰ͖Δͷ͸ ʔ

Slide 65

Slide 65 text

/ 86 eXplainable AI ʢXAIɿઆ໌ՄೳAIʣ 22 Wang, D., Yang, Q., Abdul, A., & Lim, B. Y. (2019, May). Designing theory-driven user-centric explainable AI. In Proceedings of the 2019 CHI conference on human factors in computing systems (pp. 1-15). ਓؒͷཧղ આ໌ՄೳAI

Slide 66

Slide 66 text

/ 86 XAIͱೝ஌όΠΞε 23 2022೥࣌఺Ͱର৅࿦จ37ຊͷΈʂ ·ͩ·ͩϒϧʔΦʔγϟϯʂ Bertrand, A., Belloum, R., Eagan, J. R., & Maxwell, W. (2022, July). How cognitive biases affect XAI-assisted decision-making: A systematic review. In Proceedings of the 2022 AAAI/ACM Conference on AI, Ethics, and Society (pp. 78-91). XAIʹΑͬͯ༠ൃ આ໌͕ݪҼͱͳΔ XAIͷධՁʹӨڹ͢Δ XAIʹΑͬͯܰݮ XAIͰ༠ൃ͞Εͨ ೝ஌όΠΞεΛܰݮ XAIٕज़

Slide 67

Slide 67 text

/ 86 XAIͱೝ஌όΠΞε 23 [1] Eiband, M., Buschek, D., Kremer, A., & Hussmann, H. (2019, May). The impact of placebic explanations on trust in intelligent systems. In Extended abstracts of the 2019 CHI conference on human factors in computing systems (pp. 1-6). [2] Cadario, R., Longoni, C., & Morewedge, C. K. (2021). Understanding, explaining, and utilizing medical artificial intelligence. Nature Human Behaviour, 5(12), 1636-1642. [3] van der Waa, J., Nieuwburg, E., Cremers, A., & Neerincx, M. (2021). Evaluating XAI: A comparison of rule-based and example-based explanations. Artificial Intelligence, 291, 103404. [4] Chen, V., Liao, Q. V., Wortman Vaughan, J., & Bansal, G. (2023). Understanding the role of human intuition on reliance in human-AI decision-making with explanations. Proceedings of the ACM on Human-Computer Interaction, 7(CSCW2), 1-32. [5] Mualla, Y., Tchappi, I., Kampik, T., Najjar, A., Calvaresi, D., Abbas-Turki, A., ... & Nicolle, C. (2022). The quest of parsimonious XAI: A human-agent architecture for explanation formulation. Artificial intelligence, 302, 103573. [6] T. Lombrozo, “Simplicity and probability in causal explanation,” Cognit. Psychol., vol. 55, no. 3, pp. 232–257, Nov. 2007 [7] Nguyen, H. A., Hofman, J. M., & Goldstein, D. G. (2022, April). Round numbers can sharpen cognition. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems (pp. 1-15). [8] Cao, H., Spatharioti, S. E., Goldstein, D. G., & Hofman, J. M. (2023). Comparing scalable strategies for generating numerical perspectives. arXiv preprint arXiv:2308.01535. [9] Alicioglu, G., & Sun, B. (2022). A survey of visual analytics for Explainable Artificial Intelligence methods. Computers & Graphics, 102, 502-520. 9"*ͱਓؒͷҙࢥܾఆɾߦಈͷؔ܎ هड़తݚڀ u ϓϥηϘઆ໌ɺӕͷઆ໌ ◇ ਖ਼౰͕ͩ৘ใΛ఻͍͑ͯͳ͍આ໌͸ਖ਼͍͠આ໌ͱಉ౳ʹΞϓϦΛ৴པͤ͞Δ <> ◇ ਖ਼͍͠આ໌ΑΓΘ͔ͬͨؾʹͳΔઆ໌͕ਓؒͷߦಈΛม͑Δ <> u 9"*ʹ͓͚Δઆ໌ܗࣜ ◇ ൓࣮ܕઆ໌ʹࣄྫܕઆ໌Λ௥Ճ͢Δͱɺઆ໌͕ؒҧ͍ͬͯͯ΋આಘͤ͞ΒΕΔɻ ͨͩ͠ɺλεΫύϑΥʔϚϯε͸޲্͠ͳ͍ɻ<> ◇ Ҽࢠܕઆ໌͸ҙࢥܾఆΛվળͤͣɺ"*͕ޡ͍ͬͯͯ΋"*Λ৴͡Δ <> u આ໌ͷ෼ྔ ◇ QBSTJNPOZʢγϯϓϧ͔ͭద੾ʣͳઆ໌ͷ࣮૷ <> ೝ஌Պֶ෼໺ͷ஌ݟʢ࠷΋޿ൣͰγϯϓϧʣ<> ͱໃ६͠ͳ͍ ϓϥηϘɿِༀɻༀͷ੒෼͕ೖ͍ͬͯͳ͍ݟͨ໨͚ͩͷༀɻ આ ໌ ͷ ಺ ༰ આ ໌ ͷ ද ݱ u ਺஋ ◇ ࢛ࣺޒೖؙͯ͠Ίͨ਺஋Λݟͤͨํ͕ɺͦͷ਺஋Λࢥ͍ग़͢ਫ਼౓͕ߴ͍ <> ◇ จ຺ʹԠͯ͡ద੾ͳදݱ͕มΘΔ <> u ࢹ֮දݱ <> ʮ ਓʯWTʮ ਓʯ ʮ ԯυϧʯWTʮΞϝϦΧ࿈๜༧ࢉͷ໿ʯ WTʮΞϝϦΧਓ͋ͨΓ໿ υϧʯ

Slide 68

Slide 68 text

/ 86 ৴པɺઆಘ 24 ৴ པ ʢ 5 S V T U ʣ "*ؔ࿈෼໺ ૊৫Ϛωδϝϯτ ࣾձ৺ཧֶ l ಓಙతடংʹର͢ Δظ଴ͷ͏ͪɺ૬ खͷ಺໘ʹ͋Δਓ ؒੑ΍ࣗ෼ʹର͢ Δײ৘ͳͲͷ൑அ ʹ΋ͱ͍ͮͯͳ͞ ΕΔɺ૬खͷҙਤ ʹ͍ͭͯͷظ଴ <> ͜͜ʹهࡌͨ͠ͷ͸֤෼໺ʹ͓͚Δ୅දతͳҰྫɻۃΊͯଟ༷ͳٞ࿦͕ࠞಱͱ ͍ͯ͠ΔͨΊɺݚڀจ຺͝ͱʹద੾ͳఆٛΛ࠾༻͢Δඞཁ͕͋Δɻ l ૬खͷҙਤ΍ߦಈ ʹର͢Δߠఆతͳ ظ଴ʹج͖ͮɺ੬ ऑੑΛड͚ೖΕΔ ҙࢥΛؚΉ৺ཧঢ় ଶ <> [1] https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai [2] Lee, J. D., & See, K. A. (2004). Trust in automation: Designing for appropriate reliance. Human factors, 46(1), 50-80. [3] Merritt, S. M., Heimbaugh, H., LaChapell, J., & Lee, D. (2013). I trust it, but I don’t know why: Effects of implicit attitudes toward automation on trust in an automated system. Human factors, 55(3), 520-534. [4] Jian, J. Y., Bisantz, A. M., & Drury, C. G. (2000). Foundations for an empirically determined scale of trust in automated systems. International journal of cognitive ergonomics, 4(1), 53-71. [5] Parasuraman, R., Sheridan, T. B., & Wickens, C. D. (2000). A model for types and levels of human interaction with automation. IEEE Transactions on systems, man, and cybernetics-Part A: Systems and Humans, 30(3), 286-297. [6] Yang, X. J., Schemanske, C., & Searle, C. (2023). Toward quantifying trust dynamics: How people adjust their trust after moment-to-moment interaction with automation. Human Factors, 65(5), 862-878. [7] Schaefer, K. (2013). The perception and measurement of human-robot trust. [8] Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after all: A cross-discipline view of trust. Academy of management review, 23(3), 393-404. [9] ࢁ؛ढ़உ. (1998). ৴པͷߏ଄. ౦ژେֶग़൛ձ. p.46-47 ࣗಈԽγεςϜ l ෆ࣮֬ੑͱ੬ऑ͞ ʹΑͬͯಛ௃෇͚ ΒΕΔঢ়گͰɺ͋ Δݸਓͷ໨ඪͷୡ ੒Λ͋ΔΤʔδΣ ϯτ͕खॿ͚͢Δ ͩΖ͏ͱ͍͏ଶ౓ <> ʔ ଌఆई౓ < > "* l ҎԼͷͭΛຬͨ ͢͜ͱ ᶃద๏ʢશͯͷద ༻๏ྩΛଚॏʣᶄ ྙཧతʢྙཧݪଇ ͱՁ஋؍Λଚॏʣ ᶅݎ࿚ੑʢࣾձ؀ ڥͱٕज़త؍఺Λ ߟྀʣ <>

Slide 69

Slide 69 text

/ 86 ৴པɺઆಘ 24 ৴ པ ʢ 5S V T U ʣ Ali, S., Abuhmed, T., El-Sappagh, S., Muhammad, K., Alonso-Moral, J. M., Confalonieri, R., ... & Herrera, F. (2023). Explainable Artificial Intelligence (XAI): What we know and what is left to attain Trustworthy Artificial Intelligence. Information Fusion, 99, 101805.

Slide 70

Slide 70 text

/ 86 ৴པɺઆಘ 24 આ ಘ )$*෼໺ʢDBQUPMPHZʣ ࣾձ৺ཧֶ l ਓͷߦಈɾߟ͑ํɾଶ౓Λม͑Δ͜ͱɺ ಈػ͚ͮΔɺ௥ै͢Δ [1] ύʔε΢ΣΠγϒٕज़͸ߟ͑ํ΍ ߦಈͷมԽʹண໨͢Δ <> l ଞऀͷଶ౓Λม͑Δ͜ͱ <> ଶ౓ŋŋŋର৅ʹؔ͢Δɺ޷Έ΍ධՁ తͳ൑அʹج͍ͮͨ৺ཧతͳ܏޲<> [1] Fogg, B. J. (1998, April). Captology: the study of computers as persuasive technologies. In CHI 98 Conference Summary on Human Factors in Computing Systems (p. 385). [2] Fogg, B. J. Persuasive technology: using computers to change what we think and do. Morgan Kaufmann, 2003 [3] ஑ాΒɺࣾձ৺ཧֶ ิగ൛ɺ༗൹ֳɺ2019ɺp.137-158 [4] Bai, H., Voelkel, J., Eichstaedt, J., & Willer, R. (2023). Artificial intelligence can persuade humans on political issues. "*෼໺ <> ੓ࡦʹؔ͢ΔGPT-3 or 3.5ͷੜ੒จΛಡΜͰɺਓ͕ؒઆಘ͞ΕͯҙݟΛม͑ͨ ਓ͕ؒهࡌͨ͠จষͱಉఔ౓ʹ

Slide 71

Slide 71 text

/ 86 ೝ஌ಛੑΛ౿·͑ͨػցֶशΞϧΰϦζϜ 25 u ਓؒͷ஌֮Λදݱ͢Δ߲ΛՃ͑ͨࣝผϞσϧ ࣄྫϕʔεͷઆ໌ʹجͮ͘࠷ऴతͳਓؒͷҙࢥܾఆ͕ɺ෼ྨਫ਼౓ʹج͍ͮͯ ։ൃ͞ΕͨϞσϧΛ࢖͏ͱ͖ΑΓվળ͢Δɻ Han Liu, Yizhou Tian, Chacha Chen, Shi Feng, Yuxin Chen, and Chenhao Tan. Learning Human-Compatible Representations for Case-Based Decision Support. In Proceedings of ICLR 2023. https://www.youtube.com/watch?v=QlOuWbPECqM ͦͷଞͷ࿦จ΋ؚΊͨಈըղઆ

Slide 72

Slide 72 text

/ 86 AIΛ࢖͏ਓؒͷҙࢥܾఆաఔΛཧղ͍ͨ͠ 26 ࣬ױ …ͳͲ ҙࢥܾఆ 50% ҩऀA . ݸਓɾूஂ) ü ü ( ü ü ü ( ҙࢥܾఆϞσϧ ᶃपғͷਓʑ ᶄঢ়گ …ͳͲ º ҩऀA ҙࢥܾఆաఔʹӨڹ͢Δཁૉ छྨʹ෼ྨ Hoff, K. A., & Bashir, M. (2015). Trust in automation: Integrating empirical evidence on factors that influence trust. Human factors, 57(3), 407-434.

Slide 73

Slide 73 text

/ 86 AIΛ࢖͏ਓؒͷҙࢥܾఆաఔΛཧղ͍ͨ͠ 26 ҙࢥܾఆաఔʹӨڹ͢ΔཁૉɿλεΫ u ٬؍తͩͱײ͡ΔλεΫͰ"*ͷճ౴Λ࠾༻͠΍͍͢ <> u ߟྀ͢΂͖ม਺͕ଟ͍λεΫͷ৔߹ɺ"*ͷճ౴Λ࠾༻͠΍͍͢ <> u ࣗಈԽγεςϜͷಈ࡞͕ద੾͔ݕূ࡞ۀ͕൥ࡶͳ৔߹ɺࣗಈԽόΠΞε͕ ੜ͡΍͍͢ <> ྫɿΞϥʔτ͕ɺ௨ৗ࣌ʹ͸ൃಈͤͣɺظ଴͞ΕΔͱ͖ͷΈʹൃಈ͢Δ͜ͱΛ֬ೝ͢Δ [1] Castelo, N., Bos, M. W., & Lehmann, D. R. (2019). Task-dependent algorithm aversion. Journal of Marketing Research, 56(5), 809-825. [2] Bansal, G., Nushi, B., Kamar, E., Lasecki, W. S., Weld, D. S., & Horvitz, E. (2019, October). Beyond accuracy: The role of mental models in human-AI team performance. In Proceedings of the AAAI conference on human computation and crowdsourcing (Vol. 7, No. 1, pp. 2-11). [3] Lyell, D., & Coiera, E. (2017). Automation bias and verification complexity: a systematic review. Journal of the American Medical Informatics Association, 24(2), 423-431.

Slide 74

Slide 74 text

/ 86 AIΛ࢖͏ਓؒͷҙࢥܾఆաఔΛཧղ͍ͨ͠ 26 ҙࢥܾఆաఔʹӨڹ͢Δཁૉɿҙࢥܾఆऀ ஌ͬͯಘ͢Δଌఆई౓ u ೝ஌ख़ྀੑςετʢ$PHOJUJWF3FqFDUJPOUFTUɿ$35ʣ<> u χϡϝϥγʔ ŋŋŋ਺తೳྗʢOVNCFSMJUFSBDZʣ< > u ᐆດ͞଱ੑʢUPMFSBODFPGBNCJHVJUZʣ<> ͦͷଞɺͨ͘͞Μ [1] Hoff, K. A., & Bashir, M. (2015). Trust in automation: Integrating empirical evidence on factors that influence trust. Human factors, 57(3), 407-434. [2] Himmelstein, M. (2022). Decline, adopt or compromise? A dual hurdle model for advice utilization. Journal of Mathematical Psychology, 110, 102695. [3] Frederick, S. (2005). Cognitive reflection and decision making. Journal of Economic perspectives, 19(4), 25-42. [4] Lipkus, I. M., Samsa, G., & Rimer, B. K. (2001). General performance on a numeracy scale among highly educated samples. Medical decision making, 21(1), 37-44. [5] Fagerlin, A., Zikmund-Fisher, B. J., Ubel, P. A., Jankovic, A., Derry, H. A., & Smith, D. M. (2007). Measuring numeracy without a math test: development of the Subjective Numeracy Scale. Medical Decision Making, 27(5), 672-680. [6] Weller, J. A., Dieckmann, N. F., Tusler, M., Mertz, C. K., Burns, W. J., & Peters, E. (2013). Development and testing of an abbreviated numeracy scale: A Rasch analysis approach. Journal of Behavioral Decision Making, 26(2), 198-212. [7] MacDonald, A. (1970). Revised scale for ambiguity tolerance: Reliability and validity. Psychological Reports, 26(3), 791-798. [8] Rebholz, T. R., & Hütter, M. (2023). Bayesian advice taking: Adaptive strategy selection in sequential advice seeking. [9] Yaniv, I., & Milyavsky, M. (2007). Using advice from multiple sources to revise and improve judgments. Organizational Behavior and Human Decision Processes, 103(1), 104-120. ! '% '$ &! "! # u ࣗಈԽγεςϜ΁ͷ৴པ จԽࠩɺ೥ྸɺੑผŋŋŋ͍ͣΕ΋֬ఆతͳࠩ͸ͳ͍ <> u ࣗݾத৺తॿݴׂҾ ౰ॳͷࣗ෼ͷߟ͑ͱॿݴͷؔ܎ੑʹґଘ zద౓zʹ཭ΕͨॿݴΛ࢖͏ < >

Slide 75

Slide 75 text

/ 86 AIΛ࢖͏ਓؒͷҙࢥܾఆաఔΛཧղ͍ͨ͠ 26 ҙࢥܾఆաఔʹӨڹ͢Δཁૉɿ؀ڥ u ෳ਺ͷॿݴΛ౷߹͢ΔҙࢥܾఆաఔͷϞσϧԽ< > [1] Bansal, G., Nushi, B., Kamar, E., Lasecki, W. S., Weld, D. S., & Horvitz, E. (2019, October). Beyond accuracy: The role of mental models in human-AI team performance. In Proceedings of the AAAI conference on human computation and crowdsourcing (Vol. 7, No. 1, pp. 2-11). [2] Rebholz, T. R., & Hütter, M. (2023). Bayesian advice taking: Adaptive strategy selection in sequential advice seeking [3] Yaniv, I., & Milyavsky, M. (2007). Using advice from multiple sources to revise and improve judgments. Organizational Behavior and Human Decision Processes, 103(1), 104-120. u "*ͷग़ྗͷ׆༻ํ๏ͷஞ࣍తมԽ <>

Slide 76

Slide 76 text

/ 86 ੜ෺ֶతج൫Λ୳ࡧ͢Δݚڀ 27 u ೴ਆܦՊֶ෼໺1 fMRI ◇ ֬ূόΠΞεͷਆܦֶతج൫ <> ◇ ॿݴऀ͕ਓؒͷͱ͖ͱػցͷͱ͖ͱͷҧ͍ <> u ೴ਆܦՊֶ෼໺2 ೴೾ ◇ ෆ׬શͳࣗಈԽγεςϜ΍ΞϧΰϦζϜΛ؂ࢹ͍ͯ͠Δͱ͖ͷڻ͖΍ظ଴ ͷݕग़ < > ◇ ਓؒͷ׆ಈΛ؂ࢹ͢Δͱ͖ͱࣗಈԽγεςϜΛ؂ࢹ͢Δͱ͖ͷҧ͍ <> u ࢹઢܭଌ ◇ 9"*Λ࢖͏ҩऀͷࢹઢ <> ◇ ࣗಈԽγεςϜΛݟΔස౓͕ߴ͍ͱ͖͸৴པ͍ͯ͠ͳ͍ <> [1] Kappes, A., Harvey, A. H., Lohrenz, T., Montague, P. R., & Sharot, T. (2020). Confirmation bias in the utilization of others’ opinion strength. Nature neuroscience, 23(1), 130-137. [2] Goodyear, K., Parasuraman, R., Chernyak, S., Madhavan, P., Deshpande, G., & Krueger, F. (2016). Advice taking from humans and machines: An fMRI and effective connectivity study. Frontiers in Human Neuroscience, 10, 542. [3] Akash, K., Hu, W. L., Jain, N., & Reid, T. (2018). A classification model for sensing human trust in machines using EEG and GSR. ACM Transactions on Interactive Intelligent Systems (TiiS), 8(4), 1-20. [4] De Visser, E. J., Beatty, P. J., Estepp, J. R., Kohn, S., Abubshait, A., Fedota, J. R., & McDonald, C. G. (2018). Learning from the slips of others: Neural correlates of trust in automated agents. Frontiers in human neuroscience, 12, 309. [5] Somon, B., Campagne, A., Delorme, A., & Berberian, B. (2019). Human or not human? Performance monitoring ERPs during human agent and machine supervision. NeuroImage, 186, 266-277. [6] Nagendran, M., Festor, P., Komorowski, M., Gordon, A., & Faisal, A. A. (2023, July). Eye-tracking of clinician behaviour with explainable AI decision support: a high-fidelity simulation study. In ICML 3rd Workshop on Interpretable Machine Learning in Healthcare (IMLH). [7] Hergeth, S., Lorenz, L., Vilimek, R., and Krems, J. F. (2016). Keep your scanners peeled: gaze behavior as a measure of automation trust during highly automated driving. Hum. Factors 58, 509–519.

Slide 77

Slide 77 text

/ 86 ·ͱΊ 29 Θ͔͍ͬͯΔ͜ͱɾͰ͖͍ͯΔ͜ͱ ͜Ε͔Βͷ՝୊ ਓؒ"*ڠௐܕҙࢥܾఆͷՊֶͷཱ֬ u ҙࢥܾఆաఔͷཧղ ◇ ೝ஌Պֶɾܦࡁֶɾ೴ਆܦՊֶͳͲଟ༷ͳ෼໺ʹ͓͚Δٞ࿦ ◇ ଟ෼໺ʹ·͕ͨΔ஌ݟͷ౷߹తཧղ u ҙࢥܾఆաఔͷཧղʹجͮ͘AIγεςϜ΁ͷ޻ֶతԠ༻ [1] ◇ AIͷΞϧΰϦζϜͷఏҊ ◇ AIΛ͏·͘׆༻Ͱ͖ΔγεςϜͷఏҊ [1] Gigerenzer, G. (2023). Psychological AI: Designing Algorithms Informed by Human Psychology. Perspectives on Psychological Science, 17456916231180597. u ਓؒ͸AIΛաখධՁͨ͠ΓաେධՁͨ͠Γ͢Δ u AIΛաখධՁͨ͠ΓաେධՁͨ͠Γ͢Δਓؒͷҙࢥܾఆաఔ͸ղ໌͞Ε͍ͯͳ͍ u աখධՁɾաେධՁ͞Εʹ͍͘AIγεςϜͷཁ݅΋ఆ·͍ͬͯͳ͍ 51FGIH:?24+MO-K-LNP?3H4C+0@<

Slide 78

Slide 78 text

͍͞͝ʹ

Slide 79

Slide 79 text

/ 86 ·ͱΊ 30 AI։ൃ γεςϜ࣮૷ ద੾ͳධՁ ਓؒΛཧղ͢Δ ਓؒͷೝ஌όΠΞεͷཧղ͸ɺAI։ൃɾγεςϜ࣮૷ͷج൫ͱͳΔɻ └ ݱ࣌఺Ͱطଘͷਓจࣾձܥͷݚڀ੒ՌΛ֎ૠ͢Δ͚ͩͰ͸ෆे෼ 7³® ¾º°™¢®©º¡

Slide 80

Slide 80 text

/ 86 ೝ஌όΠΞεΛθϩʹ͸Ͱ͖ͳ͍ 31 ೝ஌όΠΞε͸[1] u AIʹಛ༗ͷݱ৅Ͱ͸ͳ͍ u ҉໧తͰແҙࣝతͳ΋ͷ [1] Schwartz, R., Vassilev, A., Greene, K., Perine, L., Burt, A., & Hall, P. (2022). Towards a standard for identifying and managing bias in artificial intelligence. NIST special publication, 1270(10.6028). [2] Gigerenzer, G. (2023). Psychological AI: Designing Algorithms Informed by Human Psychology. Perspectives on Psychological Science, 17456916231180597. [3] Han Liu, Yizhou Tian, Chacha Chen, Shi Feng, Yuxin Chen, and Chenhao Tan. Learning Human-Compatible Representations for Case-Based Decision Support. In Proceedings of ICLR 2023. [4] Honda, H., Kagawa, R., & Shirasuna, M. (2023). The nature of anchor-biased estimates and its application to the wisdom of crowds. ೝ஌όΠΞεΛ׆༻͢Δ͜ͱͰɺਓؒͱAIͷڠௐతҙࢥܾఆΛΑΓྑ͍΋ͷʹ u Ғ͍ઌੜͷఏݴ [2] u ۩ମతͳݚڀྫ ◇ ػցֶशΞϧΰϦζϜ΁ͷ૊ΈࠐΈ [3] ◇ ΑΓྑ͍ू߹஌΁ͷ׆༻ [4] ೝ஌όΠΞεΛθϩʹ͸Ͱ͖ͳ͍ ೝ஌όΠΞε͸ਓؒΒ͠͞ ¢²ac˜8²t= u _SÓÂÀÌ¿H Sº°™²³b¡¨™ u t¤¨S#ž M/¥¼°º’Ç–Þ–Ò³Sª¬œŸ¨™ u %­Ÿ§š°_SÓÂÀ̳ %¥¶Ÿ’×ÌÑÖÛÈÐÁ̲p³Q3

Slide 81

Slide 81 text

/ 86 Take Home Message 32 ਺஋ʹͨ͠ॠؒʹҰਓา͖ͯ͠͠·͏΋ͷ͸ੈʹଟ͋͘Δɻ ͋͋ݴ͑͹ɺ͜͏ݴ͏ɻ