על התופעה מרחיבים לדוגמה במאמר הזה https://t.co/b8yJMCii5J
RT @AltePute: никогда не представлено всё общество. В эпоху социальных сетей этот эффект искусственно усиливается, формируется информационн…
@soodoku This modern classic used URL text for a hard news classifier https://t.co/gIaNZmunij
@cphoffmann @ajungherr To make matters worse, some of the existing canon of literature is confounded by similar COIs and problems related to research ethics/standards. @facebook coauthors and data curation were a problem back in 2015. https://t.co/ASOdEOlE
@Leesplez Also check out: https://t.co/nq4SlUmxEy
Exposure to ideologically diverse news and opinion on Facebook | Science https://t.co/JXOh1t2Bu9
@theactualspider do you think the algorithms are in large part a function of the consumer? https://t.co/46Hwgp2sa6 find that individual choice plays a larger role in selective exposure to cross cutting information than algorithms
This study is a prime example of "LaTeX-driven advertising", ie. a scientific text used to cater to a company's marketing operations. Like the infamous 2015 Facebook article that exonerated FB's algorithm by blaming users for the creation of echo chambers.
@GaelVaroquaux @cristiancanton @ylecun What are you looking for exactly? CDS has worked on lots of social impact or social dynamics questions. Pulling some random examples: https://t.co/DKtEqnMv8q https://t.co/3RTjeWfFmD https://t.co/WMrdgBQmfL
I had another project with w/ brilliant co-authors @eytan & @ladamic, which ended up as a chapter in my dissertation & eventually became this paper https://t.co/KaUkK9wQXG.
никогда не представлено всё общество. В эпоху социальных сетей этот эффект искусственно усиливается, формируется информационный пузырь: в фейсбуке алгоритм настройки ленты новостей и круга общения изолирует людей от мнений, противоречащих их собственным. h
@ylelkes @shandlin Total brain fart on my behalf for forgetting this classic. 2500 citations in 6 years https://t.co/gILoVEiIks
@stephanetrainel @Aurelie_JEAN Ce qui m'ennuie c'est qu'a priori il y a autant d'études qui disent que les bulles de filtre n'existent pas ou ne sont pas aussi pire qu'on peut le penser. Par ex: https://t.co/JCLu2tJHKO Le fait de croire sans étude nous con
@CalebTCarrPhD This article is good...but I think the graphics are blue and red...it's behind a paywall and I'm on my phone: https://t.co/1pM1Hc2y3E
Then ties this into what we know about Russian bots from House Intelligence reports! Really interesting. @HZinnbrooks shares some research on co-followership networks that helps put this in a math framework reference: https://t.co/GoXWG3UGOI
RT @acerbialberto: …or that around 30 percent of the political content shared in Facebook by one’s friends is of the opposite ideology” (th…
RT @grinbergnir: @kombiz Essentially yes. The calculation is detailed in Bakshy, Messing & Adamic's paper from 2015 (link below) and we're…
@kombiz Essentially yes. The calculation is detailed in Bakshy, Messing & Adamic's paper from 2015 (link below) and we're using their scores. https://t.co/5Dr6Dh00Wf
For example, see: https://t.co/xTSjU1ZgA3 https://t.co/LiYHva9T4q https://t.co/4owSVpNVYQ [3/n]
i co teraz? ;) https://t.co/5j3u3jPBxT @widownia https://t.co/qiMxepthed
Today is a good day to retweet this. - First, what @SolomonMg said. - Second, "Compared with algorithmic ranking, individuals’ choices played a stronger role in limiting exposure to cross-cutting content." - Third, great paper - https://t.co/UmMt4Gpr79
https://t.co/Fflp4pOYmo 池谷先生が以前つぶやかれていた「人間はインターネットを見る際に自分に必要な情報にしか注意を払わない」という論文 これによって、生きづらい人たちの非機能的な認知がより強くなっていく側面はあるんじゃないかと思う。
@Tatsuya_S_Zuki So many ways to approach this problem. With so much hidden behind proprietary algorithms. It is both a base-rate issue (remember: https://t.co/1pM1Hck9se) and an algorithmic misalignment.
@alexstamos @brandonsilverm @rowlsmanthorpe @kevinroose Yes, definitely. And certainly work by FB scientists has worked with both URL canonicalization and aggregations by publisher eg https://t.co/qUV47x5W46 https://t.co/i9W6cEErCc
Excelente
@sebacampanario @uadlup @aracalacana Esta es una referencia clave, que dice que elegimos nuestra red por afinidad antes de que los algoritmos jueguen un rol: https://t.co/SaKlZ1vcas Igual mi heurística para todo tema político es "lo que diga Ezra". Nunca
RT @carlosdiuk: @santiagollach @sebacampanario @joaconavajas @ProgramaAEDuns @uadlup @danyscht Hay estudios contradictorios sobre el rol de…
RT @carlosdiuk: @santiagollach @sebacampanario @joaconavajas @ProgramaAEDuns @uadlup @danyscht Hay estudios contradictorios sobre el rol de…
RT @carlosdiuk: @santiagollach @sebacampanario @joaconavajas @ProgramaAEDuns @uadlup @danyscht Hay estudios contradictorios sobre el rol de…
@santiagollach @sebacampanario @joaconavajas @ProgramaAEDuns @uadlup @danyscht Hay estudios contradictorios sobre el rol de redes sociales en polarización. Acá hay dos bien directamente en dirección opuesta a la idea de Kakutani: https://t.co/eWE6TncojF o
@derSchett Zu Algorithmen kann ich außerdem nur nochmals auf folgende wissenschaftliche Studie verweisen. Und damit da kein Missverständnis entsteht, bin bei allem ganz bei dir. https://t.co/A2mEdJDYpI
…or that around 30 percent of the political content shared in Facebook by one’s friends is of the opposite ideology” (this proportion goes down a bit because of algorithmic filtering) 5/8 https://t.co/LWKPnafqYR
RT @JaMoEberl: Forschung, die es gerade nicht mehr in den Beitrag geschafft hat... z.B. https://t.co/Ywe7fn144w Zur Fremdsteuerung durch A…
Forschung, die es gerade nicht mehr in den Beitrag geschafft hat... z.B. https://t.co/Ywe7fn144w Zur Fremdsteuerung durch Algorithmen auf Facebook: "Compared with algorithmic ranking, individuals’ choices played a stronger role in limiting exposure to cro
@jorgejhms Filter bubble: cuando el algoritmo te hace la selección de contenidos similares a tu ideología, reforzándola Cámara de eco: cuando solo escuchas o dialogas con gente que tiene una ideología similar a la tuya, pero reforzándola https://t.co/OUpD
y a ces 2 liens vers des études scientifiques sur la page wikipédia, qui semble s'opposer à la théorie du Filter Bubble : est ce qu'il y aurait d'autres études qui les contrediraient, et ou qui seraient davantage reconnues ? https://t.co/IEvTO5qTmp https
@byjacobward Well, Facebook published one a few years ago. https://t.co/HC1KN3soNs Lots of responses/caveats followed.
@HenrikOscarsson En sökning på Google Scholar tycks någorlunda samstämmig med Bergström när det gäller rekommendationer från Google News, men kanske ej när det gäller Facebook. https://t.co/XCWoRXkENO
@gdfm7 But there's plenty more. Did you read this: https://t.co/mqK5gvCsBJ? I'd argue otherwise. What's the super robust evidence for the existence of echo chambers and filter bubbles?
@TurkerNational @tristanharris Now, I read a lot of your posts because I found what you were Tweeting about interesting. I like what this study shows us as we consider a pure anti-tech anti-algorithm approach: https://t.co/O7Se4tnEfh 2/2
@pdakean @LindaSkitka I think a growing consensus is that social media does contribute to political polarization, but relationship is complex & likely due more to social factors than exposure. see: https://t.co/so9IRQc614, https://t.co/csyOnB4zyX, http
the paper: https://t.co/1SARbOBnvS
@rayovirtual "FB tiene el algoritmo que te junta solo con tus similares, pero Twitter no." Twitter también hace feed ranking, no hay diferencia fundamental en cómo funcionan. Y en ambos casos es más complejo y el efecto más fuerte es la auto selección. htt
@brodnig @NikoAlm Btw. hier der Link zur FB-Studie: https://t.co/aWppgHejoM
#neverstopimproving #NeverStopLearning The study: Exposure to ideologically diverse news and opinion on Facebook https://t.co/glUoq02M8G The book that the line was taken from: Never Stop Learning: Stay Relevant, Reinvent Yourself, and Thrive @brstaats @
人天性趋向“择群而居”,而社交媒体和推荐算法更是加剧这一偏好。观点相左的两派人看到对方侧的新闻和评论的机会被进一步缩小。 https://t.co/5Fa4BiIJEg
@d_a_robson @BBC_Future Hey David, even Facebook-funded research confirms the echo chamber: https://t.co/ge9qaXkekX Twitter also facilitates selective-exposure: https://t.co/Qd80Hy7sjy https://t.co/H6MPGGQjma Here is an explanation for the reason: ht
2/8 "Exposure to ideologically diverse news and opinion on Facebook" https://t.co/147HeapXTi https://t.co/S80c8U0I6Q
People’s choices are more important than algorithms in limiting exposure ideologically discordant content on Facebook. https://t.co/ATlE4MhxKz 8/14 https://t.co/vE2BXJr0SU
RT @yuji_ikegaya: 「多様さ」は受けられない?インターネットには様々な情報が飛び交っているのにヒトは特定の情報にしか注意を払わないそうです。今朝の『サイエンス』論文→http://t.co/ycPtgOm2Ui(閲覧履歴から嗜好を推測して推奨する自動アルゴリズム…
RT @davidthewid: @oatesmeal Social media bubbles: https://t.co/0oHpEphZhz https://t.co/TwcS3qR2UG Polarization: https://t.co/c6xyHMzsyk
@oatesmeal Social media bubbles: https://t.co/0oHpEphZhz https://t.co/TwcS3qR2UG Polarization: https://t.co/c6xyHMzsyk
@jayvanbavel Last quibble: Lit on differences between libs/cons on social media is not consistent. E.g., Facebook data doesn't seem to show that, if anything it shows the opposite. https://t.co/wwloLpVfrX https://t.co/lsMw1WULbT
@ThatDamnYank @jburnmurdoch Except there isn't any evidence that likes and RTs create a reinforcing bubble. For example, this study by Facebook shows that echo chambers don't really exist. https://t.co/HzKh0iWvkI
@McFaul Research on this question makes me think it mostly reflects pre-existing divides rather than causing new ones: https://t.co/Ue9wl7XXyc https://t.co/ZVM6PnZI2f https://t.co/h3B0ptz6Ub
RT @yuji_ikegaya: 「多様さ」は受けられない?インターネットには様々な情報が飛び交っているのにヒトは特定の情報にしか注意を払わないそうです。今朝の『サイエンス』論文→http://t.co/ycPtgOm2Ui(閲覧履歴から嗜好を推測して推奨する自動アルゴリズム…
Exposure to ideologically diverse news and opinion on #Facebook https://t.co/if7zeraa52 #BigData #AI
@bdunogier @gchampeau sure. there is this one : https://t.co/rTZs15Vl7H and also https://t.co/27SCGYbJG9
Exposure to ideologically diverse news and opinion on Facebook https://t.co/8WqmXeBMpa
@spangsege @dragz Originalidea from Pariser,study in Science found FB algorithms contributed to differences but less than partisanship https://t.co/l1mlFi0mdx
Estudos quantitativos para EGC. https://t.co/ahofVuyWFz
@dinapomeranz @ThewissenS A more recent study finding the opposite effect: https://t.co/2BMWKwOHXX
@gunthert Bon OK la prochaine fois, je lis les notes de bas de page avant de poser une question https://t.co/rZUMxPjU4p ;-)
RT @dominic_egger: Gibt es #Filterblasen oder nicht? Ja, gibt es, aber sie ist klein: https://t.co/affihddSGt Diese einzige Studie kommt vo…
Gibt es #Filterblasen oder nicht? Ja, gibt es, aber sie ist klein: https://t.co/affihddSGt Diese einzige Studie kommt von #Facebook. #rp17
@buddy_hell Not peer reviewed =/= Not worth reading. But for peer reviewed see https://t.co/aTjZaN28tT or https://t.co/5HczOXgoKr
RT @yuji_ikegaya: 「多様さ」は受けられない?インターネットには様々な情報が飛び交っているのにヒトは特定の情報にしか注意を払わないそうです。今朝の『サイエンス』論文→http://t.co/ycPtgOm2Ui(閲覧履歴から嗜好を推測して推奨する自動アルゴリズム…
Can fake news on Facebook really influence users? https://t.co/8qZDPZvEEc #FakeNewsMedia
Favourite from 2015: "Exposure to ideologically diverse news and opinion on Facebook": https://t.co/F6xT9BqHYn. #Facebook
RT @ExtraAI: Favourite from 2015: "Exposure to ideologically diverse news and opinion on Facebook": https://t.co/Q2A09b67xl. Next few tweet…
Favourite from 2015: "Exposure to ideologically diverse news and opinion on Facebook": https://t.co/Q2A09b67xl. Next few tweets from there.
Exposure to ideologically diverse news and opinion on Facebook https://t.co/ITHe23AUoH
Exposure to ideologically diverse news and opinion on Facebook https://t.co/UpKYgk3pui
RT @ecalvo68: @christilcara @piscitelli @edmundogs @rguareschi @cscolari @aracalacana @ugemitch https://t.co/04H0XHA8MY Una version mas lar…
RT @ecalvo68: @christilcara @piscitelli @edmundogs @rguareschi @cscolari @aracalacana @ugemitch https://t.co/04H0XHA8MY Una version mas lar…
RT @ecalvo68: @christilcara @piscitelli @edmundogs @rguareschi @cscolari @aracalacana @ugemitch https://t.co/04H0XHA8MY Una version mas lar…
@christilcara @piscitelli @edmundogs @rguareschi @cscolari @aracalacana @ugemitch https://t.co/04H0XHA8MY Una version mas larga.
Exposure to ideologically diverse news and opinion on Facebook | Science https://t.co/ElZ45d9gki via @instapaper
RT @DrMJoyner: Political science. Exposure to ideologically diverse news and opinion on Facebook. - PubMed - NCBI https://t.co/UzIlxi7I8d
When assessing social and cultural phenomena data>beliefs and anecdotes https://t.co/PebNL44ByX
RT @DrMJoyner: Political science. Exposure to ideologically diverse news and opinion on Facebook. - PubMed - NCBI https://t.co/UzIlxi7I8d
Political science. Exposure to ideologically diverse news and opinion on Facebook. - PubMed - NCBI https://t.co/UzIlxi7I8d
RT @lamethodeFC: Parution mai 2015 dans #Science > Exposure to ideologically diverse news and opinion on #Facebook https://t.co/0dSDWtslGq…
Parution mai 2015 dans #Science > Exposure to ideologically diverse news and opinion on #Facebook https://t.co/0dSDWtslGq #LaMethSci https://t.co/9VPA4hjkLI
あら、そういえば日本の大学で”political science” をうたっているところはある? 法学部政治学科か経済学部政治学科しか見たことない。共産党かな史的唯物論 https://t.co/KfuOtNXSVC
Exposure to ideologically diverse news and opinion on Facebook | Science https://t.co/8numVuxUX9
Exposure to ideologically diverse news and opinion on Facebook https://t.co/bmLmvvWr3r
@Jalonen @KetonenOksi @mihkal @jnkka @jjussila @AramoHeli @ElinaLappalaine WSJ on hieno! Mutta klikkaukset myös. Ks. https://t.co/giCc4fGhuA
This 2015 research by Facebook's own team showed a (pretty small) reduction in exposure to different perspectives https://t.co/aIGC9bnoiw https://t.co/ipb9cdppBf
RT @BasHofstra: . @StukRoodVlees Zie bijvoorbeeld hier: https://t.co/zDcjZUhMBa en hier: https://t.co/yf08UtFdp4 en hier: https://t.co/eZpG…
. @StukRoodVlees Zie bijvoorbeeld hier: https://t.co/zDcjZUhMBa en hier: https://t.co/yf08UtFdp4 en hier: https://t.co/eZpGJAbZYO
Exposure to ideologically diverse news and opinion on Facebook. https://t.co/ptw1GmjB5a
@davidsirota It's based on this: https://t.co/UdfFY1aqVC written by this guy: https://t.co/N3t6YmCL2s. It's facebook statistics.
@Autumm Facebook actually found it was people's behavior, rather than their algorithm. https://t.co/EkZxj0C5Cn
.@mkeulemans "klopt dat wel?" n.a.v column @Petermiddendorp? Filterbubble valt namelijk mee: https://t.co/xiiiv6UswX https://t.co/eOezuINNpW
@deaneckles idk. the sources seem totally disjoint. but your take agrees w Bakshy et al https://t.co/RvsYxtwSsJ https://t.co/6RIiFdUpTp
I've talked about this a lot this week so retweeting: Exposure to ideologically diverse news and opinion on Facebook https://t.co/YMT1gKegKQ
L'esposizione alle #notizie e il confronto delle #idee nell'epoca di #Facebook #PhD_FB https://t.co/VJakPpgjnE …