サイバー空間におけるフェイクニュースの広がりとその対策

Slide 1

Slide 1 text

αΠόʔۭؒʹ͓͚ΔϑΣΠΫχϡʔε ͷ޿͕Γͱͦͷରࡦ ଜࢁଠҰ ԣ඿ࠃཱେֶ ॿڭ .BJMNVSBZBNBUBJDIJCT!ZOVBDKQ ICT-ISAC ⼤阪WG-Day

Slide 2

Slide 2 text

2 ൃද಺༰ എܠ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ ੜ੒"*ͱϑΣΠΫχϡʔε 4/4ΛऔΓר͘؀ڥʢΞςϯγϣϯɾΤίϊϛʔʣ ࠷ۙͷऔΓ૊Έʢӄ๳࿦ͱಡॻʣ

Slide 3

Slide 3 text

Slide 4

Slide 4 text

5 ϑΣΠΫχϡʔεͷ൙ཞ എܠ બڍલϲ݄Ͱ໿ສճɺީิऀਓ τϥϯϓࢯ ΫϦϯτϯࢯ ͷϑΣΠΫχϡʔε͕'BDFCPPLͰ֦ࢄ Allcott, Hunt, and Matthew Gentzkow. "Social media and fake news in the 2016 election." Journal of economic perspectives 31.2 (2017): 211-36.

Slide 5

Slide 5 text

6 ϑΣΠΫχϡʔεͷ൙ཞ എܠ 5G network spreads COVID-19 Vitamin C cures COVID-19

Slide 6

Slide 6 text

7 ϑΣΠΫχϡʔε࡞੒ͷϝϦοτ എܠ https://www.newsweek.com/2017/06/16/big-data-mines-personal-info-manipulate-voters-623131.html ੈ࿦ૢ࡞ ͓ۚṶ͚ https://www.nbcnews.com/news/world/fake-news-how-partying-macedonian-teen- earns-thousands-publishing-lies-n692451

Slide 7

Slide 7 text

8 ϑΣΠΫχϡʔεͷඃ֐ എܠ ๫ಈࣄ݅ ͓ۚṶ͚ https://www.npr.org/2018/07/18/629731693/fake-news-turns-deadly-in-india https://www.washingtonpost.com/news/worldviews/wp/2013/04/23/syrian-hackers-claim-ap-hack-that- tipped-stock-market-by-136-billion-is-it-terrorism/ גՁٸམ

Slide 8

Slide 8 text

9 ϑΣΠΫχϡʔεͷඃ֐ɿ੓࣏ എܠ ⼭⼝真⼀（2022）『ソーシャルメディア解体全書』、勁草書房 l ϑΣΠΫχϡʔεʹΑͬͯର৅ ΁ͷࢧ࣋Λ௿Լͤ͞ΔޮՌ l ಛʹऑ͍ࢧ࣋Λࢯ͍ͯͨਓʑʹ ର͠ϑΣΠΫχϡʔε͸ޮՌత ʹӨڹΛ༩͑Δ

Slide 9

Slide 9 text

10 ϑΣΠΫχϡʔεΛආ͚Δࢦ਑ͷ࡞੒ എܠ https://www.weforum.org/agenda/2020/11/tackling-covid-19-misinformation-who/

Slide 10

Slide 10 text

11 ൃද಺༰ എܠ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ ੜ੒"*ͱϑΣΠΫχϡʔε 4/4ΛऔΓר͘؀ڥʢΞςϯγϣϯɾΤίϊϛʔʣ ࠷ۙͷऔΓ૊Έʢӄ๳࿦ͱಡॻʣ

Slide 11

Slide 11 text

12 ιʔγϟϧϝσΟΞ͸৘ใ֦ࢄ૷ஔ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε

Slide 12

Slide 12 text

13 l ೔ຊͰ͸ϛυϧϝσΟΞͷଘࡏ͕ιʔγϟϧϝσΟΞͰͷϑΣΠ Ϋχϡʔε֦ࢄͷҰ໾Λ୲͍ͬͯΔͱ͍͏ࢦఠ https://www.soumu.go.jp/main_content/000749423.pdf ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε

Slide 13

Slide 13 text

14 ϑΣΠΫχϡʔεͷ֦ࢄ͸଎ͨ͘͘͞Μ఻ΘΔ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε Vosoughi, Soroush, Deb Roy, and Sinan Aral. "The spread of true and false news online." science 359.6380 (2018): 1146-1151. https://www.soumu.go.jp/main_content/000630427.pdf l ಛʹ੓࣏ɺϏδωεɺઓ ૪ɺՊֶͳͲͷ࿩୊Ͱ֦ ࢄ͞Ε΍͍͢܏޲ l #PUͷޮՌ͸ҙ֎ͱখ͍͞

Slide 14

Slide 14 text

15 ϑΣΠΫχϡʔεΛ֦ࢄ͢ΔϢʔβͱ͸ʁ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε Yamaguchi, Shinichi, and Tsukasa Tanihara. "Relationship between misinformation spreading behaviour and true/false judgments and literacy: an empirical analysis of COVID-19 vaccine and political misinformation in Japan." Global Knowledge, Memory and Communication (2023). ౰વ͕ͩɺͦͷχϡʔεΛʮਖ਼͍͠ʯͱ ৴ͨ͡ΓϝσΟΞɾϦςϥγʔ͕௿͍ਓ ͕֦ࢄ͠қ͍ ϑΣΠΫχϡʔεͷڞ༗ͷ͸ͷ ϢʔβʹΑΔ΋ͷͰɺߴྸऀ΍ӈدΓͷ ੓࣏ࢥ૝ͷϢʔβ͕ڞ༗͠΍͍͢ Grinberg, Nir, et al. "Fake news on Twitter during the 2016 US presidential election." Science 363.6425 (2019): 374-378.

Slide 15

Slide 15 text

16 ϑΣΠΫχϡʔεΛݕग़͢ΔࢼΈ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε l ͲͷΑ͏ͳಛ௃Λ༻͍Δͷ͔ʁ l ༻͍ͨಛ௃Λ্ख͘׆༻Ͱ͖Δ͔ػցֶशϞσϧΛͲͷΑ͏ʹ ߏங͢Δͷ͔ʁ l ͲͷΑ͏ͳ෇ՃՁ஋ΛՃ͑Δͷ͔ʁʢ৴པੑͳͲʣ 機械学習モデルフェイクニュースの可能性のある投稿に関する情報 (特徴量) ⼊⼒出⼒ Fake or True

Slide 16

Slide 16 text

17 ϑΣΠΫχϡʔεΛݕग़Ϟσϧͷྫ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε l ςΩετ৘ใ΍ڞ༗ͨ͠Ϣʔ βͷಛ௃Λ༻͍ͯɺϑΣΠΫ χϡʔε͔Ͳ͏͔Λ൑ఆ l ൑ఆʹ஫໨ͨ͠ՕॴΛՄࢹԽ ʢઆ໌ੑͷ෇༩ʣ Lu, Yi-Ju, and Cheng-Te Li. "GCAN: Graph-aware Co-Attention Networks for Explainable Fake News Detection on Social Media." Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 2020.

Slide 17

Slide 17 text

18 ϑΣΠΫχϡʔεΛݕग़Ϟσϧͷ՝୊ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε l ೥ͷσʔλͰߏஙͨ͠ϑΣΠΫχϡʔεݕग़Ϟσϧ͕೥ ͷσʔλʹରԠͰ͖Δͷ͔ʁ EJBDISPOJDCJBTͷଘࡏ l ϞσϧΛຖ೥ߋ৽͠ଓ͚Δͱ͍͏ͷ͸େม l ߏங͞ΕΔػցֶशϞσϧͷ൑அ͸ਖ਼͍͠Θ͚Ͱ͸ͳ͍ l ਖ਼͍͠৘ใΛϑΣΠΫͱࢦఠ͢Δةݥੑʹ͍ͭͯ ِཅੑ΍ِӄੑ l ػցֶशϞσϧͷϒϥοΫϘοΫε໰୊ l ൑அͷࠜڌ΍࢓૊Έ͕Θ͔Βͳ͍ͱਓؒ͸ͦͷϞσϧΛ৴͖͡Εͳ͍ l େྔʹଘࡏ͢Δ౤ߘΛͲͷΑ͏ʹࡹ͔͘

Slide 18

Slide 18 text

19 (VBSEJBOͷ׆༻ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε l (VBSEJBOϑΣΠΫχϡʔεΛࢦఠ͢Δ Ϣʔβ l (VBSEJBOΛࣗಈతʹൃݟɺੜ੒͢Δ͜ ͱͰϑΝΫτνΣοΫίετͷ࡟ݮ l గਖ਼Ϣʔβ͕ެࣜΞΧ΢ϯτͳͲͰ͋Δ ͱޡใͷࢦఠΛ৴͡ΔՄೳੑ͕Ξοϓ Hannak, Aniko, et al. "Get back! you don’t know me like that: The social mediation of fact checking interventions in twitter conversations." Eighth International AAAI Conference on Weblogs and Social Media. 2014. Vo, Nguyen, and Kyumin Lee. "Learning from fact-checkers: analysis and generation of fact-checking language." Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. 2019. Seo, Haeseung, et al. "If You Have a Reliable Source, Say Something: Effects of Correction Comments on COVID-19 Misinformation." Proceedings of the International AAAI Conference on Web and Social Media. Vol. 16. 2022.

Slide 19

Slide 19 text

20 ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ϑΣΠΫχϡʔεͷ֦ࢄɿ೔ຊͷྫ https://www.asahi.com/articles/ASN2X6CXLN2XULFA03L.html https://www.asahi.com/articles/ASQ5D7X35Q5BULEI00R.html

Slide 20

Slide 20 text

21 ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ʮ$07*%ʹΑͬͯτΠϨοτϖʔύʔ͕඼ബ͸ޡΓʯ ͱ͍͏౤ߘ͕ສҎ্ͷ35

Slide 21

Slide 21 text

22 ྫτΠϨοτϖʔύʔ૽ಈJO$07*% ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ೥݄೔͔Β೥݄೔ͷτΠϨοτϖʔύʔ͕ ؚΉ UXFFUT

Slide 22

Slide 22 text

23 ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε

Slide 23

Slide 23 text

24 ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε $07*%ʹΑΔτΠϨοτϖʔύʔ඼੾Ε౤ߘ 35ҎԼ࡟আࡁΈ

Slide 24

Slide 24 text

25 ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε τΠϨοτϖʔύʔͷ඼੾Εͷใࠂ 21. より詳細な分析: ⿃海不⼆夫, 榊剛史, and 吉⽥光男. "ソーシャルメディアを⽤いた新型コロナ禍における感情変化の分析." ⼈⼯知能学会論⽂誌 35.4 (2020): F-K45_1.

Slide 25

Slide 25 text

26 l ೔ຊͰ͸ϛυϧϝσΟΞͷଘࡏ͕ιʔγϟϧϝσΟΞͰͷϑΣΠ Ϋχϡʔε֦ࢄͷҰ໾Λ୲͍ͬͯΔͱ͍͏ࢦఠ https://www.soumu.go.jp/main_content/000749423.pdf ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε

Slide 26

Slide 26 text

27 ൃද಺༰ എܠ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ ੜ੒"*ͱϑΣΠΫχϡʔε 4/4ΛऔΓר͘؀ڥʢΞςϯγϣϯɾΤίϊϛʔʣ ࠷ۙͷऔΓ૊Έʢӄ๳࿦ͱಡॻʣ

Slide 27

Slide 27 text

28 ਓʑͷ࣋ͭೝ஌όΠΞε ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ l ଟ͘ͷਓʑ͸ܦݧ΍௚ײ͔ Βޡͬͨ൑அΛͯ͠͠·͏ l ͜ΕΒͷόΠΞε͕ਓʑΛ ϑΣΠΫχϡʔεΛ৴͡͞ ͤΔޮՌΛ΋ͨΒ͢ҰํͰ ɺϑΣΠΫχϡʔεରࡦʹ ࢖͑ΔՄೳੑ΋ൿΊ͍ͯΔ https://en.wikipedia.org/wiki/Cognitive_bias

Slide 28

Slide 28 text

29 όοΫϑΝΠΞʔޮՌ ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ l ݸਓͷ৴೦ʹ߹Θͳ͍గਖ਼Λ༩͑ΒΕͨ ͱͯ͠΋ɺͦͷగਖ਼Λ৴ͣ͡ɺ͔͑ͬͯ గਖ਼ʹΑͬͯݸਓͷ৴೦ΛڧΊΔޮՌ l όοΫϑΝΠΞޮՌΛࣔ͢Ϣʔβ͸ɺෆ ฏෆຬ΍εϥϯάతͳදݱΛ༻͍Δ͜ͱ ͕ଟ͍ Nyhan, Brendan, and Jason Reifler. "When corrections fail: The persistence of political misperceptions." Political Behavior 32.2 (2010): 303-330. Jiang, Shan, and Christo Wilson. "Linguistic signals under misinformation and fact-checking: Evidence from user comments on social media." Proceedings of the ACM on Human-Computer Interaction 2.CSCW (2018): 1-23.

Slide 29

Slide 29 text

30 "WBJMBCJMJUZ)FVSJTUJD ʢར༻ՄೳੑώϡʔϦεςΟοΫʣ ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ l ࠷ۙ໨ʹͨ͠৘ใ΍සൟʹ઀৮͢Δ৘ใ ͕൑அʹӨڹΛ༩͑Δ͜ͱ l ϝσΟΞ΍4/4Ͱසൟʹ໨ʹ͢ΔϑΣΠ Ϋχϡʔε͸ɺޡͬͨ৘ใ͕ࣄ࣮ͱͯ͠ ೝࣝ͞Ε΍͍͢ Pennycook, Gordon, et al. "Shifting attention to accuracy can reduce misinformation online." Nature 592.7855 (2021): 590-595.

Slide 30

Slide 30 text

31 ΫΠζΛ׆༻ͨ͠ϑΣΠΫχϡʔε΁ͷରࡦ ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ l ϑΣΠΫχϡʔεΛݟ͔ͯΒగਖ਼͢ΔͷͰ ͸ͳ͘ɺʮڭҭʯͱͯ͠ϑΣΠΫχϡʔε ʹର͢ΔϫΫνϯΛ઀छ l ΫΠζܗࣜͰֶͿ͜ͱͰɺϑΣΠΫχϡʔ εʹରͯ͠ὃ͞Εʹ͘͘ͳΔޮՌΛൃش Roozenbeek, Jon, and Sander van der Linden. "Breaking Harmony Square: A game that “inoculates” against political misinformation." The Harvard Kennedy School Misinformation Review (2020).

Slide 31

Slide 31 text

32 ಈըΛ׆༻ͨ͠ϑΣΠΫχϡʔε΁ͷରࡦ ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ l ಈըʹΑΔϑΣΠΫχϡʔεରࡦ l ϑΣΠΫχϡʔεʹରͯ͠ڧ͘ͳΔʮ৺ ཧత༧๷઀छʯͷޮՌ͸ݟΒΕ͕ͨɺΫ Πζ΄Ͳͷେ͖͞Ͱ͸ͳ͍ l ҰํͰɺίετ͕҆͘ɺଟ͘ͷਓʑʹ ʮडಈతʯʹ઀छ͕Մೳ Roozenbeek, Jon, et al. "Psychological inoculation improves resilience against misinformation on social media." Science advances 8.34 (2022): eabo6254.

Slide 32

Slide 32 text

33 φοδΛ׆༻ͨ͠ϑΣΠΫχϡʔε΁ͷରࡦ ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ l χϡʔεΛݟΔલʹ5SVF'BLFϥϕϧΛఏ ࣔ͢Δ΂͖͔ʁݟͨޙʹఏࣔ͢Δ΂͖͔ɺ ʹ͍ͭͯͲͷλΠϛϯάͰఏࣔ͢Δ͔Ͱ ਓʑͷϑΣΠΫχϡʔεʹର͢Δೝ஌͕ม Խ͢Δ͔Λݕূ l ݟग़͠ΛಡΜͩޙʹగਖ਼͢Δ͜ͱ͕Ұ൪ޮ Ռ͕ߴ͍͜ͱΛࣔ͢ Brashier, Nadia M., et al. "Timing matters when correcting fake news." Proceedings of the National Academy of Sciences 118.5 (2021): e2020043118.

Slide 33

Slide 33 text

34 ൃද಺༰ എܠ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ ੜ੒"*ͱϑΣΠΫχϡʔε 4/4ΛऔΓר͘؀ڥʢΞςϯγϣϯɾΤίϊϛʔʣ ࠷ۙͷऔΓ૊Έʢӄ๳࿦ͱಡॻʣ

Slide 34

Slide 34 text

35 ੜ੒"*ͷൃల ੜ੒"*ͱϑΣΠΫχϡʔε

Slide 35

Slide 35 text

36 ੜ੒"*ʹΑΔ%FFQGBLFͷ࡞੒ ੜ੒"*ͱϑΣΠΫχϡʔε from: https://gulfbusiness.com/deepfakes-novel-trend-or-novel-threat/ l %FFQGBLF%FFQ-FBSOJOHͱ GBLFΛ૊Έ߹Θ͋Θͤͨ଄ޠ l Πϯλʔωοτ্Ͱ؍࡯͞Εͨ %FFQGBLFͷ਺͸૿Ճ܏޲

Slide 36

Slide 36 text

37 'BLF/FXT(FOFSBUJPO ੜ੒"*ͱϑΣΠΫχϡʔε Brown, Tom, et al. "Language models are few-shot learners." Advances in neural information processing systems 33 (2020): 1877-1901. (15ʹΑΔχϡʔεੜ੒

Slide 37

Slide 37 text

38 'BLF/FXT(FOFSBUJPO ੜ੒"*ͱϑΣΠΫχϡʔε https://news.yahoo.co.jp/expert/articles/846ed869620be2126f3cb749df6139d7282eb975 l Ұ໨Ͱ͸ݟ෼͚͕͔ͭͳ͍ը૾ ੜ੒ʹΑΔσϚ ྫɿz4IJ[VPLBEJTBTUFSGMPPEz IUUQTTUBCMFEJGGVTJPO XFCDPNQMBZHSPVOE

Slide 38

Slide 38 text

39 ੜ੒"*ͷ੓࣏৘ใ΁ͷ׆༻ l ੓࣏ɾࣾձ໰୊ʹؔ͢Δ৘ใ Λ࡞੒͢ΔͨΊʹɺը૾ɾς Ωετͷੜ੒ϞσϧΛͷࠃ Ͱར༻͞Ε͍ͯΔ͜ͱΛใࠂ House, Freedom. The repressive power of artificial intelligence. Technical Report.{Freedom House}. https://freedomhouse. org/report/freedom-net/2023/repressive-power-artificial-intelligence, 2023. ੜ੒"*ͱϑΣΠΫχϡʔε

Slide 39

Slide 39 text

40 ੜ੒"*ʹΑΔ࡞੒͞Εͨχϡʔεͷઆಘྗ l (15͸ਓؒΑΓ΋આಘྗͷ͋Δِ৘ใΛ࡞Γग़͢ l ਓؒʹ͸χϡʔε͕ੜ੒"*ʹΑΔ΋ͷ͔Ͳ͏͔൑அ Ͱ͖ͳ͍ Spitale, Giovanni, Nikola Biller-Andorno, and Federico Germani. "AI model GPT-3 (dis) informs us better than humans." Science Advances 9.26 (2023): eadh1850. ੜ੒"*ͱϑΣΠΫχϡʔε

Slide 40

Slide 40 text

41 ϑΣΠΫχϡʔεݕग़ʹ--.͕࢖͑Δ͔ʁ Li, Guanghua, et al. "Re-Search for The Truth: Multi-round Retrieval-augmented Large Language Models are Strong Fake News Detectors." arXiv preprint arXiv:2403.09747 (2024). ੜ੒"*ͱϑΣΠΫχϡʔε l --.ͱݕࡧΛ૊Έ߹ΘͤΔ͜ͱͰχϡʔε͕'BLF͔Ͳ͏͔Λ൑ఆ

Slide 41

Slide 41 text

42 %FFQGBLFݕग़ٕज़ͷൃల Reiss, Tal, Bar Cavia, and Yedid Hoshen. "Detecting deepfakes without seeing any." arXiv preprint arXiv:2311.01458 (2023). ੜ੒"*ͱϑΣΠΫχϡʔε https://www.i.u-tokyo.ac.jp/news/press/2022/202204262039.shtml

Slide 42

Slide 42 text

43 ੜ੒"*࣌୅ͷϑΣΠΫχϡʔε ੜ੒"*ͱϑΣΠΫχϡʔε l ϑΣΠΫχϡʔεͷߴ౓ԽɾେྔԽ l ΠϯϑΥΧϦϓεͷ౸དྷʢʁʣ l ϩγΞɾ΢ΫϥΠφઓ૪͕ࣔ͢Α͏ʹɺ৘ใઓ͕ॏཁͳཱͪҐஔͱ ͳΔ l ϓϥοτϑΥʔϜࣄۀऀͷ໾ׂ͕ॏཁͱͳΔ

Slide 43

Slide 43 text

44 ൃද಺༰ എܠ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ ੜ੒"*ͱϑΣΠΫχϡʔε 4/4ΛऔΓר͘؀ڥͱ·ͱΊ ࠷ۙͷऔΓ૊Έʢӄ๳࿦ͱಡॻʣ

Slide 44

Slide 44 text

45 ΤίʔνΣϯόʔݱ৅ l 4/4্Ͱɺࣗ෼ͱࣅͨΑ͏ͳՁ஋؍΍ ߟ͑ํͷϢʔβʔΛϑΥϩʔ͢Δ͜ͱ Ͱɺಉ͡Α͏ͳχϡʔε΍৘ใ͹͔Γ ͕ྲྀ௨͢Δดͨ͡৘ใ؀ڥ l ෼ۃԽ΍ϑΣΠΫχϡʔεຮԆͷԹচ 4/4ΛऔΓר͘؀ڥ

Slide 45

Slide 45 text

46 ೥݄ͷΞϝϦΧͷ4/4ۭؒ 4/4ΛऔΓר͘؀ڥ Nikolov, Dimitar, Alessandro Flammini, and Filippo Menczer. "Right and left, partisanship predicts (asymmetric) vulnerability to misinformation." arXiv preprint arXiv:2010.01462 (2020).

Slide 46

Slide 46 text

47 ϑΟϧλʔόϒϧ l Ϣʔβͷݸਓ৘ใΛֶशͨ͠ΞϧΰϦζϜʹΑͬͯɺͦͷਓʹͱͬ ͯڵຯؔ৺͕͋Γͦ͏ͳ৘ใ͹͔Γ͕΍ͬͯ͘Δ৘ใ؀ڥ l (PPHMFݕࡧ΍4/4ͳͲ΄ͱΜͲͷαʔϏεʹඋ͑ΒΕ͍ͯΔਪન γεςϜͷӨڹʢύʔιφϥΠθʔγϣϯٕज़ͷෛͷଆ໘ʣ 4/4ΛऔΓר͘؀ڥ https://www.soumu.go.jp/main_content/000630427.pdf

Slide 47

Slide 47 text

48 ΞςϯγϣϯɾΤίϊϛʔ l ϓϥοτϑΥʔϜͷ୆಄ʹΑ ΓΞςϯγϣϯࢢ৔ͷ୆಄ l ਓʑͷ࣌ؒΛୣ͍औΔڝ૪ۭ ؒͷܗ੒ l “Most people are unaware of how they spend their time and attention, their most crucial resources.” 4/4ΛऔΓר͘؀ڥ https://www.marketingmag.com.au/tech-data/why-attention-is-the-worlds-most-valuable- resource/

Slide 48

Slide 48 text

49 ΞςϯγϣϯɾΤίϊϛʔ l Ξςϯγϣϯࢢ৔Ͱ͸ʮࢥ૝ͷڝ૪ʯ͔Βʮܹࢗͷڝ૪ʯ΁ l ϑΣΠΫχϡʔε͕૿͑Δཧ༝ͷͭ l 5SPMMͷ૿ՃͷཁҼʹ΋ l "*ͷ੒௕͸ਓʑͷ࠷దͳ"UUFOUJPO֫ಘํ๏Λ༧ଌ͢Δ l 5JL5PLͷॎεΫϩʔϧʹΑΔਪન l γεςϜʹΑͬͯਓʑͷڧ੍తͳʮ൓ࣹʯΛҾ͖ى͜͢͜ͱ͔Βɺ ʮनΘΕͷ௃ऩʢDBQUJWFBVEJFODFʣʯͱදݱ l ೔ຊͰ͸ిं಺ͷԻ੠޿ࠂ͕ਓ֨ݖ֐ʹ͋ͨΔ͔Ͱٞ࿦͕Ճ೤ͨ͜͠ͱ΋ 4/4ΛऔΓר͘؀ڥ

Slide 49

Slide 49 text

50 ΞςϯγϣϯɾΤίϊϛʔ l 5JL5PLͷϨίϝϯυγ εςϜͷෆಁ໌ੑ΍ະ ੒೥΁ͷϦεΫ͔Βσ δλϧαʔϏε๏ͷٛ ຿ҧ൓ͷ͍͕͔͚ٙΒ ΕΔʢ&6ʣ 4/4ΛऔΓר͘؀ڥ https://www.asahi.com/articles/ASS2M7TT6S2MUHBI02B.html

Slide 50

Slide 50 text

51 ࠓޙͷσδλϧ؀ڥ l ϓϥοτϑΥʔϜͷ໾ׂ͕·͢·͢ॏཁʹͳ͍ͬͯ͘ l Ͳ͜·Ͱن੍͠ɺͲ͜·Ͱදݱͷࣗ༝Λҡ࣋͢Δ؀ڥͱ͢Δͷ͔ʁ l ֎ࢿܥͷϓϥοτϑΥʔϜࣄۀऀ͕ଟ͍தͰɺϢʔβʹ೔ຊޠͰରԠͰ͖ Δମ੍Λ੔͑Δ͔ʁ l %FFQ'BLFɺϝλόʔεͳͲͷٕज़ͷֵ৽ʹ൐͍ɺੜ͡Δࣾձͷ ໰୊΋ෳࡶԽ l ৘ใϦςϥγʔڭҭͷ֦ॆ ͕ٸ຿ σδλϧ؀ڥͷະདྷ

Slide 51

Slide 51 text

52 ࠓޙͷσδλϧ؀ڥ l ϑΝΫτνΣοΫ΍ίϛϡχςΟϊʔτͳͲͷऔΓ૊Έͷ֦ॆͷ ඞཁੑ l ೔ຊͰ͸ϑΝΫτνΣοΫػ͕ؔগͳ͍ʢϑΝΫτνΣοΫͷଟ༷ੑ͕อ ͨΕͳ͍ʣ l ϚεϝσΟΞ͕ͦͷ໾ׂΛ ୲͏͜ͱ͕ظ଴͞Ε͍ͯΔ σδλϧ؀ڥͷະདྷ ⼭⼝ほか（2022）「Innovation Nippon 2021 わが国における偽・誤情報の実態の把握と社会的対処の検討」 https://www.glocom.ac.jp/activities/project/7759

Slide 52

Slide 52 text

53 ࠓޙͷσδλϧ؀ڥ ੍౓ɺγεςϜʢφοδʣɺϦςϥγʔͷͭΛಉ࣌ʹ ڧ͍ͯ͘͘͜͠ͱ͕ॏཁ l ೔ຊ͸ωοτ؀ڥ΍σδλϧίϯςϯπʹର͢Δ๏੍౓͕຀Վత l ϦςϥγʔڭҭɺϓϨόϯΩϯάͷऔΓ૊ΈΛૣΊʹऔΓ૊ΜͰ͍͘͜ͱ͕ॏ ཁ σδλϧ؀ڥͷະདྷ

Slide 53

Slide 53 text

54 ൃද಺༰ എܠ ιʔγϟϧϝσΟΞͱϑΣΠΫχϡʔε ͳͥਓʑ͸ϑΣΠΫχϡʔεΛ৴ͯ͡͠·͏ͷ͔ʁ ੜ੒"*ͱϑΣΠΫχϡʔε 4/4ΛऔΓר͘؀ڥͱ·ͱΊ ࠷ۙͷऔΓ૊Έʢӄ๳࿦ͱಡॻʣ

Slide 54

Slide 54 text

56 ಡॻཤྺ͔Βӄ๳࿦܏౗ϢʔβΛൃݟΛ໨ࢦ͢ ࠷ۙͷऔΓ૊Έ l ಡॻه࿥αʔϏεɿಡॻϝʔλʔ l ֤Ϣʔβ͕ʮ͍ͭʯɺʮԿͷຊʯΛͲͷΑ͏ͳײ૝Λ࣋ͬͯಡΜ͔ͩه࿥ l ӄ๳࿦ʹؔ͢ΔຊΛಡΜͩϢʔβͷཤྺ͕ݟΕΔ

Slide 55

Slide 55 text

57 ಡॻཤྺσʔλͷ෼ੳ ࠷ۙͷऔΓ૊Έ l ܭࢉػՊֶతΞϓϩʔνΛద༻Մೳͳܗʹม׵ 埋め込み表現技術ベクトル表現 0.1, 0.3, 0.4, …, 0.9, 0.2, 0.4 0.2, 0.1, 0.3, …, 0.1, 0.0, 0.2 0.3, 0.5, 0.1, …, 0.0, 0.3, 0.5 Book1 Book2 Book3 Book4 Book1 Book2 Book3

Slide 56

Slide 56 text

58 ຊͷҐஔ͚ͮͷՄࢹԽ ࠷ۙͷऔΓ૊Έ

Slide 57

Slide 57 text

59 ࠷ۙͷऔΓ૊Έ ॻ੶ͷຒΊࠐΈදݱ

Slide 58

Slide 58 text

60 ࠷ۙͷऔΓ૊Έ 32ӄ๳࿦܏౗ಡऀʢλʔήοτϢʔβ܈ʣ͸ͲͷΑ͏ͳॻ੶Λ޷ΜͰಡΉ܏ ޲ʹ͋ΔͷͩΖ͏͔ʁ ωοτϫʔΫ ΫϥελϦϯά ࣌ؒܦա աڈͷಡॻཤྺ λʔήοτϢʔβ λʔήοτϢʔβ܈͕ӄ๳࿦ॻ੶ʹग़ձ͏·ͰʹಡΜͰ͖ͨຊ͔Βɼଞͷ Ϣʔβͱൺֱͯ͠ಛʹಡ·Ε͍ͯΔॻ੶ΫϥεΛݟ͚ͭΔ

Slide 59

Slide 59 text

61 ࠷ۙͷऔΓ૊Έ 32ӄ๳࿦܏౗ಡऀʢλʔήοτϢʔβ܈ʣ͸ͲͷΑ͏ͳॻ੶Λ޷ΜͰಡΉ܏ ޲ʹ͋ΔͷͩΖ͏͔ʁ

Slide 60

Slide 60 text

62 ࠷ۙͷऔΓ૊Έ 32ӄ๳࿦܏౗ಡऀʢλʔήοτϢʔβ܈ʣ͸ͲͷΑ͏ͳॻ੶Λ޷ΜͰಡΉ܏ ޲ʹ͋ΔͷͩΖ͏͔ʁ

Slide 61

Slide 61 text

63 ࠷ۙͷऔΓ૊Έ 32ӄ๳࿦܏౗ಡऀʢλʔήοτϢʔβ܈ʣ͸ͲͷΑ͏ͳॻ੶Λ޷ΜͰಡΉ܏ ޲ʹ͋ΔͷͩΖ͏͔ʁ

Slide 62

Slide 62 text

64 ࠷ۙͷऔΓ૊Έ 32ӄ๳࿦܏౗ಡऀʢλʔήοτϢʔβ܈ʣ͸ͲͷΑ͏ͳॻ੶Λ޷ΜͰಡΉ܏ ޲ʹ͋ΔͷͩΖ͏͔ʁ

Slide 63

Slide 63 text

65 ࠷ۙͷऔΓ૊Έ 32ӄ๳࿦܏౗ಡऀʢλʔήοτϢʔβ܈ʣ͸ͲͷΑ͏ͳॻ੶Λ޷ΜͰಡΉ܏ ޲ʹ͋ΔͷͩΖ͏͔ʁ