{"id":22069,"date":"2018-03-20T00:05:35","date_gmt":"2018-03-19T21:05:35","guid":{"rendered":"https:\/\/bilimvegelecek.com.tr\/?p=22069"},"modified":"2018-03-20T00:29:03","modified_gmt":"2018-03-19T21:29:03","slug":"twitterda-yanlis-bilgiler-dogru-bilgilerden-daha-hizli-yayiliyor","status":"publish","type":"post","link":"https:\/\/bilimvegelecek.com.tr\/index.php\/2018\/03\/20\/twitterda-yanlis-bilgiler-dogru-bilgilerden-daha-hizli-yayiliyor","title":{"rendered":"Twitter\u2019da yanl\u0131\u015f bilgiler, do\u011fru bilgilerden daha h\u0131zl\u0131 yay\u0131l\u0131yor"},"content":{"rendered":"

Massachusetts Teknoloji Enstit\u00fcs\u00fc\u2019nden \u00fc\u00e7 akademisyen taraf\u0131ndan yap\u0131lan yeni bir \u00e7al\u0131\u015fma, sosyal payla\u015f\u0131m platformu Twitter\u2019da yanl\u0131\u015f haberlerin do\u011fru haberlerden daha h\u0131zl\u0131 bir \u015fekilde yay\u0131ld\u0131\u011f\u0131n\u0131 ve g\u00f6rece daha b\u00fcy\u00fck oranda payla\u015f\u0131ld\u0131\u011f\u0131n\u0131 ortaya koydu. Dahas\u0131, yanl\u0131\u015f haberlerin yay\u0131lmas\u0131, yanl\u0131\u015f hik\u00e2yeleri yaymak i\u00e7in programlanm\u0131\u015f olan botlardan kaynaklanm\u0131yor. Bu haberlerin yay\u0131lma h\u0131z\u0131n\u0131n fazla olmas\u0131n\u0131n sebebi, yanl\u0131\u015f \u00f6\u011feler i\u00e7eren haberleri retweetleyen insanlar.<\/em><\/p>\n

Makalenin yazarlar\u0131ndan, MIT Sloan \u0130\u015fletme Okulu\u2019ndan Prof. Sinan Aral, \u201cGer\u00e7ek d\u0131\u015f\u0131 olan\u0131n, t\u00fcm bilgi kategorilerinde \u00f6nem derecesine ba\u011fl\u0131 olarak ger\u00e7eklerden daha h\u0131zl\u0131, daha derin bi\u00e7imde ve daha uza\u011fa yay\u0131ld\u0131\u011f\u0131n\u0131 bulduk\u201d ifadelerini kullan\u0131yor.<\/p>\n

Yine \u00e7al\u0131\u015fman\u0131n yazarlar\u0131ndan olan, Media Lab\u2019\u0131n Sosyal Makineler Laboratuvar\u0131\u2019nda (The Media Lab\u2019s Laboratory for Social Machines, LSM) y\u00f6netici ve MIT Medya Laboratuvar\u0131\u2019nda Medya Sanatlar\u0131 ve Bilimleri profes\u00f6r\u00fc olan Deb Roy \u201cBu bulgular, \u00e7evrimi\u00e7i ileti\u015fim ekosistemimizin temel y\u00f6nlerine yeni bir \u0131\u015f\u0131k tutuyor\u201d diyor. Roy, ara\u015ft\u0131rmac\u0131lar\u0131n Twitter\u2019daki yanl\u0131\u015f ve ger\u00e7ek haberler aras\u0131ndaki ili\u015fki kar\u015f\u0131s\u0131nda \u201c\u015fa\u015f\u0131rmak ile afallamak aras\u0131nda bir yerde\u201d olduklar\u0131n\u0131 da ekliyor.<\/p>\n

\"\"
Foto\u011fraftakiler (soldan sa\u011fa): Media Lab\u2019\u0131n Sosyal Makineler Laboratuar\u0131\u2019nda doktora sonras\u0131 program\u0131nda Soroush Vosoughi (oturan), MIT Sloan \u0130\u015fletme Okulu\u2019ndan David Austin Profes\u00f6r\u00fc Sinan Aral ve 2013-2017 aras\u0131 Twitter\u2019\u0131n Ba\u015f Medya Bilimcisi olarak g\u00f6rev yapan, MIT Media Lab\u2019da medya sanatlar\u0131 ve bilimleri profes\u00f6r\u00fc olan Deb Roy.<\/figcaption><\/figure>\n

Dahas\u0131 ara\u015ft\u0131rmac\u0131lar, yanl\u0131\u015f bilginin yay\u0131lmas\u0131n\u0131n esasen yanl\u0131\u015f hik\u00e2yeleri yaymaya programlanan botlardan kaynaklanmad\u0131\u011f\u0131n\u0131 ortaya koyuyor. Bunun yerine, yanl\u0131\u015f haberler, yanl\u0131\u015f \u00f6\u011feler i\u00e7eren haberleri retweetleyen ki\u015filer nedeniyle Twitter\u2019da daha h\u0131zl\u0131 yay\u0131l\u0131yor.<\/p>\n

Doktora ara\u015ft\u0131rmas\u0131yla \u00e7al\u0131\u015fman\u0131n geli\u015fmesine yard\u0131mc\u0131 olan, makalenin yard\u0131mc\u0131 yazarlar\u0131ndan, ayn\u0131 zamanda LSM\u2019de doktora sonras\u0131 \u00e7al\u0131\u015fmas\u0131 i\u00e7in bulunan Soroush Vosoughi \u015f\u00f6yle diyor: \u201cVeri k\u00fcmemizdeki t\u00fcm botlar\u0131 kald\u0131rd\u0131\u011f\u0131m\u0131zda, yanl\u0131\u015f ve do\u011fru haberlerin yay\u0131lmas\u0131 aras\u0131ndaki farklar ortaya \u00e7\u0131kt\u0131.\u201d<\/p>\n

Yap\u0131lan \u00e7al\u0131\u015fma bu durumu \u00f6l\u00e7menin \u00e7e\u015fitli yollar\u0131n\u0131 sunuyor: \u00d6rne\u011fin, yanl\u0131\u015f haber hik\u00e2yelerinin retweetlenmi\u015f olma olas\u0131l\u0131\u011f\u0131, ger\u00e7ek hik\u00e2yelerden y\u00fczde 70 oran\u0131nda daha y\u00fcksek. Ayn\u0131 zamanda ger\u00e7ek haberlerin 1500 ki\u015fiye ula\u015fmas\u0131, yanl\u0131\u015f haberlere g\u00f6re alt\u0131 kat daha fazla zaman almakta. Twitter\u2019daki k\u0131r\u0131lmam\u0131\u015f retweet zincirleri s\u00f6z konusu oldu\u011funda, yalanlar ger\u00e7eklerden yakla\u015f\u0131k 10-20 kat daha h\u0131zl\u0131 zincir derinli\u011fine ula\u015fmakta ve yanl\u0131\u015f bilgiler her bir farkl\u0131 kullan\u0131c\u0131 taraf\u0131ndan, her zincir derinli\u011finde, do\u011fru bilgilerden daha yayg\u0131n bi\u00e7imde retweetlenmekte.<\/p>\n

\u00c7al\u0131\u015fmaya dair \u201cDo\u011fru ve Yanl\u0131\u015f Haberin \u00c7evrimi\u00e7i Yay\u0131l\u0131m\u0131\u201d ba\u015fl\u0131kl\u0131 makale, ge\u00e7ti\u011fimiz g\u00fcnlerde Science<\/em> dergisinde yay\u0131mland\u0131.<\/p>\n

Acayiplik, neden yanl\u0131\u015f\u0131n yay\u0131l\u0131m\u0131n\u0131 y\u00f6nlendiriyor?<\/strong><\/p>\n

\u00c7al\u0131\u015fman\u0131n \u00e7\u0131k\u0131\u015f noktalar\u0131ndan biri, Twitter\u2019da b\u00fcy\u00fck bir ilgi g\u00f6ren 2013 Boston Maratonu bombalamalar\u0131 ve ard\u0131ndan gelen kay\u0131plar.<\/p>\n

Vosoughi, \u201cTwitter ana haber kayna\u011f\u0131m\u0131z oldu\u201d diyor, fakat trajik olaylar\u0131n ard\u0131ndan, \u201cfark ettim ki, sosyal medyada okudu\u011fum \u015feyin \u00f6nemli bir par\u00e7as\u0131 s\u00f6ylentilerdi ve bunlar ayn\u0131 zamanda yanl\u0131\u015f haberlerdi\u201d diye ekliyor.<\/p>\n

Vosoughi, o zamanki y\u00fcksek lisans dan\u0131\u015fman\u0131 Roy ile birlikte, doktora konusunu de\u011fi\u015ftirmeye ve Twitter\u2019daki s\u00f6ylentilerin do\u011frulu\u011funu tahmin edebilecek bir model geli\u015ftirme \u00fczerinde \u00e7al\u0131\u015fmas\u0131na karar verdiler.<\/p>\n

Yine Vosoughi\u2019nin lisans\u00fcst\u00fc dan\u0131\u015fmanlar\u0131ndan biri olan ve sosyal a\u011flar\u0131 kapsaml\u0131 olarak inceleyen Aral ile yapt\u0131klar\u0131 g\u00f6r\u00fc\u015fme sonras\u0131, \u00fc\u00e7 ara\u015ft\u0131rmac\u0131 yeni \u00e7al\u0131\u015fmada deneyecekleri yakla\u015f\u0131m\u0131 belirlediler: Twitter \u00fczerinden \u00e7al\u0131\u015fma i\u00e7in bir yol haritas\u0131 \u00e7\u0131karmak ve haber hik\u00e2yelerini nesnel olarak do\u011fru veya yanl\u0131\u015f olarak tan\u0131mlamak.<\/p>\n

Twitter da bu ara\u015ft\u0131rmaya destek sa\u011flad\u0131 ve MIT ekibine, ar\u015fivlerine tam eri\u015fim hakk\u0131 verdi. Roy, 2013\u2019ten 2017\u2019ye kadar Twitter\u2019\u0131n ba\u015f medya bilimcisi olarak g\u00f6rev yapt\u0131.<\/p>\n

\u00c7al\u0131\u015fma kapsam\u0131nda ara\u015ft\u0131rmac\u0131lar, 2006\u2019dan 2017\u2019ye kadar yakla\u015f\u0131k 3 milyon ki\u015fi taraf\u0131ndan toplamda yakla\u015f\u0131k 4,5 milyon kez tweetlenen, \u00f6nemli bir yay\u0131lma alan\u0131na sahip yakla\u015f\u0131k 126.000 basamakl\u0131 haber hik\u00e2yesini takip ettiler.<\/p>\n

Ara\u015ft\u0131rma ekibi, hik\u00e2yelerin do\u011fru mu yanl\u0131\u015f m\u0131 oldu\u011funu belirlemek i\u00e7in, alt\u0131 do\u011fruluk kontrol\u00fc kurulu\u015funun de\u011ferlendirmelerini kulland\u0131 (factcheck.org, hoax-slayer.com, politifact.com, snopes.org, truthorfiction.com ve urbanlegends.about.com) ve bu de\u011ferlendirmelerin, kendi kararlar\u0131yla y\u00fczde 95\u2019ten fazla bir oranda \u00f6rt\u00fc\u015ft\u00fc\u011f\u00fcn\u00fc tespit ettiler.<\/p>\n

\u0130ncelenen 126.000 tweet zincirinde, 45.000 civar\u0131ndaki haberle en b\u00fcy\u00fck kategori politikayd\u0131; onu \u015fehir efsaneleri, ticaret, ter\u00f6rizm, bilim, e\u011flence ve do\u011fal afetler gibi kategoriler izliyordu. Yanl\u0131\u015f haberlerin yay\u0131lmas\u0131, di\u011fer kategorilerdeki haberlerden ziyade politika haberleri i\u00e7in daha belirgindi.<\/p>\n

Ara\u015ft\u0131rmac\u0131lar ayn\u0131 zamanda, ara\u015ft\u0131rman\u0131n konusu olarak \u201cyanl\u0131\u015f haber\u201d terimini, \u00e7ok geni\u015f anlamlar i\u00e7eren ve her yerde kar\u015f\u0131la\u015f\u0131labilen bir niteli\u011fe sahip olan \u201csahte haber\u201d teriminden ayr\u0131 olarak ele ald\u0131lar.<\/p>\n

\u00c7al\u0131\u015fman\u0131n sonu\u00e7 b\u00f6l\u00fcm\u00fcnde yer alan bulgular temel bir soru ortaya koyuyor: Twitter\u2019da ger\u00e7ek olmayan bilgiler neden hakikatten daha h\u0131zl\u0131 yay\u0131l\u0131yor? Aral, Roy ve Vosoughi, cevab\u0131n insan psikolojisinde olabilece\u011fini \u00f6ne s\u00fcr\u00fcyor: \u201cAcayip (\u00e7.n. al\u0131\u015f\u0131lmad\u0131k olan) \u015feyleri seviyoruz.\u201d<\/p>\n

Aral, \u201cYanl\u0131\u015f haber daha al\u0131\u015f\u0131lmad\u0131k ve insanlar\u0131n al\u0131\u015f\u0131lmad\u0131k bilgileri payla\u015fma olas\u0131l\u0131klar\u0131 daha y\u00fcksek\u201d diyor. Ve sosyal a\u011flarda, insanlar \u00f6nceden bilinmeyen (ama muhtemelen yanl\u0131\u015f) bilgileri ilk payla\u015fan ki\u015fi olarak dikkat \u00e7ekebilir. B\u00f6ylece, Aral\u2019\u0131n dedi\u011fi gibi, \u201c(Do\u011frulu\u011fa bak\u0131lmaks\u0131z\u0131n) al\u0131\u015f\u0131lmad\u0131k bilgileri payla\u015fan insanlar, \u00e7o\u011fu ki\u015finin bilmedi\u011fi bilgilere sahip ki\u015filer olarak g\u00f6r\u00fcl\u00fcyor.\u201d<\/p>\n

MIT akademisyenleri, ara\u015ft\u0131rmalar\u0131nda, yanl\u0131\u015f hik\u00e2yeleri ortaya atan ve yay\u0131lmas\u0131na neden olan Twitter kullan\u0131c\u0131lar\u0131 aras\u0131ndan rasgele se\u00e7im yaparak ve se\u00e7ilenlerin yayd\u0131klar\u0131 hik\u00e2yelere verilen reaksiyonlar\u0131n i\u00e7eri\u011fini analiz ederek bu \u201cal\u0131\u015f\u0131lmad\u0131kl\u0131k hipotezini\u201d inceliyorlar.<\/p>\n

Sonu\u00e7? Vosoughi, \u201cYanl\u0131\u015f haberler ve ger\u00e7ek haberler i\u00e7in farkl\u0131 duygusal profiller g\u00f6rd\u00fck\u201d diyor. Ger\u00e7ek \u00f6yk\u00fclerin yaratt\u0131\u011f\u0131 cevaplar daha \u00e7ok genel olarak \u00fcz\u00fcnt\u00fc, beklenti ve g\u00fcven ile karakterize olurken \u201c\u0130nsanlar yanl\u0131\u015f haberlere daha fazla \u015fa\u015fk\u0131nl\u0131k ve nefretle tepki veriyor\u201d notunu ekliyor.<\/p>\n

Ara\u015ft\u0131rmac\u0131lar, \u201cacayipli\u011fin retweetlere kendi kendine yol a\u00e7t\u0131\u011f\u0131n\u0131 iddia edememelerine\u201d ra\u011fmen, makalede s\u00f6yledikleri gibi, insanlar\u0131n, ger\u00e7ek d\u0131\u015f\u0131l\u0131\u011f\u0131n acayipli\u011fi fikriyle uyumlu yanl\u0131\u015f haberler g\u00f6rd\u00fcklerinde, onun yay\u0131lmas\u0131n\u0131n \u00f6nemli bir par\u00e7as\u0131 olabildiklerini ortaya koyuyorlar.<\/p>\n

\"\"\u0130leri ara\u015ft\u0131rmalar i\u00e7in fikirler<\/strong><\/p>\n

\u00dc\u00e7 ara\u015ft\u0131rmac\u0131n\u0131n \u00fc\u00e7\u00fc de, bulgular\u0131ndaki etkinin b\u00fcy\u00fckl\u00fc\u011f\u00fcn\u00fcn son derece anlaml\u0131 oldu\u011funu d\u00fc\u015f\u00fcn\u00fcrken, bu bulgular\u0131n sosyal ya\u015fam \u00fczerine etkisiyle ilgili g\u00f6r\u00fc\u015fleri birbirinden az da olsa farkl\u0131l\u0131k g\u00f6steriyor. Aral, bulgular\u0131n toplumsal ya\u015fam a\u00e7\u0131s\u0131ndan \u201c\u00e7ok korkutucu\u201d oldu\u011funu s\u00f6ylerken, Roy konuya biraz daha iyimser yakla\u015f\u0131yor. Ancak ara\u015ft\u0131rmac\u0131lar, yanl\u0131\u015f bilginin yay\u0131lmas\u0131n\u0131 s\u0131n\u0131rlaman\u0131n yollar\u0131n\u0131 d\u00fc\u015f\u00fcnmenin \u00f6nemli oldu\u011fu \u00fczerinde ortakla\u015f\u0131yorlar ve bulgular\u0131n\u0131n konuyla ilgili daha fazla ara\u015ft\u0131rmay\u0131 te\u015fvik edece\u011fini umuyorlar.<\/p>\n

Aral, \u201c\u015eimdi yanl\u0131\u015f haberlerin yay\u0131lmas\u0131n\u0131 durdurmak i\u00e7in verdi\u011fimiz m\u00fccadelede davran\u0131\u015fsal m\u00fcdahaleler daha da \u00f6nemli hale geliyor\u201d diyor. \u201cSadece botlar bu duruma neden olsayd\u0131, teknolojik bir \u00e7\u00f6z\u00fcme ihtiyac\u0131m\u0131z olurdu.\u201d<\/p>\n

Vosoughi ise, baz\u0131 insanlar kas\u0131tl\u0131 olarak yanl\u0131\u015f haberleri yay\u0131yorken di\u011ferleri bunu bilmeden yap\u0131yorsa, bu durumda sorunun birden fazla taktik gerektirebilecek iki ayr\u0131 problemle ifade edilebilece\u011fini \u00f6ne s\u00fcr\u00fcyor. Roy, bulgular\u0131n sosyal a\u011flar, reklamc\u0131lar ve di\u011fer taraflar i\u00e7in \u201ckalite testine d\u00f6n\u00fc\u015febilecek \u00f6l\u00e7\u00fctler veya g\u00f6stergeler\u201d olu\u015fturmaya yard\u0131mc\u0131 olabilece\u011fini s\u00f6yl\u00fcyor.<\/p>\n

MIT ara\u015ft\u0131rmac\u0131lar\u0131, ayn\u0131 fenomenin Facebook dahil olmak \u00fczere di\u011fer sosyal medya platformlar\u0131nda ortaya \u00e7\u0131kmas\u0131n\u0131n da m\u00fcmk\u00fcn oldu\u011funu, ancak bu vb. konular i\u00e7in dikkatli \u00e7al\u0131\u015fmalara ihtiya\u00e7 duyuldu\u011funu vurguluyorlar. Ayn\u0131 do\u011frultuda Aral, \u201cBilimin daha fazla \u00e7al\u0131\u015fma yapabilmek i\u00e7in hem end\u00fcstri alan\u0131ndan hem de devletten daha fazla deste\u011fe sahip olmas\u0131 gerekti\u011fini\u201d s\u00f6yl\u00fcyor. Roy ya\u015fad\u0131\u011f\u0131m\u0131z an i\u00e7in \u015funlar\u0131 vurguluyor, iyi niyetli her Twitter kullan\u0131c\u0131s\u0131 basit bir fikri \u00fczerine kafa yorabilir: \u201cRetweetlemeden \u00f6nce d\u00fc\u015f\u00fcn\u00fcn.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":"

Massachusetts Teknoloji Enstit\u00fcs\u00fc\u2019nden \u00fc\u00e7 akademisyen taraf\u0131ndan yap\u0131lan yeni bir \u00e7al\u0131\u015fma, sosyal payla\u015f\u0131m platformu Twitter\u2019da yanl\u0131\u015f haberlerin do\u011fru haberlerden daha h\u0131zl\u0131 bir \u015fekilde yay\u0131ld\u0131\u011f\u0131n\u0131 ve g\u00f6rece daha b\u00fcy\u00fck oranda payla\u015f\u0131ld\u0131\u011f\u0131n\u0131 ortaya koydu. Dahas\u0131, yanl\u0131\u015f haberlerin yay\u0131lmas\u0131, yanl\u0131\u015f hik\u00e2yeleri yaymak i\u00e7in programlanm\u0131\u015f olan botlardan kaynaklanm\u0131yor. Bu haberlerin yay\u0131lma h\u0131z\u0131n\u0131n fazla olmas\u0131n\u0131n sebebi, yanl\u0131\u015f \u00f6\u011feler i\u00e7eren haberleri retweetleyen […]<\/p>\n","protected":false},"author":864,"featured_media":22072,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[19,2686,31],"tags":[1948,998,999,2685,2684,2683],"class_list":["post-22069","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-bilim-gundemi","category-iletisim-bilimleri","category-teknoloji","tag-habercilik","tag-sosyal-medya","tag-twitter","tag-yanlis-bilgi","tag-yanlis-haberler","tag-yeni-medya"],"acf":[],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/posts\/22069","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/users\/864"}],"replies":[{"embeddable":true,"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/comments?post=22069"}],"version-history":[{"count":0,"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/posts\/22069\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/media\/22072"}],"wp:attachment":[{"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/media?parent=22069"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/categories?post=22069"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/bilimvegelecek.com.tr\/index.php\/wp-json\/wp\/v2\/tags?post=22069"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}