Zimene muyenera kuwerenga kenako

• Chiyambi (gawo 2.1)

Mtundu umodzi wa kuwona zomwe sizinaphatikizidwe mu chaputala ichi ndi ethnography. Kuti mudziwe zambiri pa ethnography m'malo ojambulira, onani Boellstorff et al. (2012) , ndi zina zambiri pa ethnography mu malo osakanikirana ndi digito, onani Lane (2016) .

• Deta yaikulu (gawo 2.2)

Palibe chigwirizano chimodzi chokha cha "deta yaikulu," koma matanthauzo ambiri akuwoneka kuti akuyang'ana pa "3 Vs": voliyumu, zosiyana, ndi kuthamanga (mwachitsanzo, Japec et al. (2015) ). Onani De Mauro et al. (2015) kuti mudziwe ndemanga.

Kuphatikizidwa kwanga ndi chiwerengero cha boma cha deta mu gawo la deta yaikulu ndizosazolowereka, ngakhale ena awonanso mlanduwu, kuphatikizapo Legewie (2015) , Connelly et al. (2016) , ndi Einav and Levin (2014) . Kuti mudziwe zamtengo wapatali za deta za boma za kafukufuku, onani Card et al. (2010) , Adminstrative Data Taskforce (2012) , ndi Grusky, Smeeding, and Snipp (2015) .

Kuwona kafukufuku waubusa kuchokera mkati mwa dongosolo la ziwerengero za boma, makamaka US Census Bureau, onani Jarmin and O'Hara (2016) . Kuti mupeze chithandizo cha kutalika kwa bukhu la kafukufuku wa kafukufuku wotsatira ku Statistics Sweden, onani Wallgren and Wallgren (2007) .

Mutu uno, ndakhala ndikuyerekeza mwachidule kafukufuku wamakhalidwe monga General Social Survey (GSS) ndi chitukuko cha deta monga Twitter. Kuti muwone bwino komanso mwatsatanetsatane pakati pa kafukufuku wamakono ndi deta, onani Schober et al. (2016) .

• Zomwe zimafanana ndi deta yaikulu (gawo 2.3)

Makhalidwe khumi awa a deta yaikulu afotokozedwa m'njira zosiyanasiyana mosiyana ndi olemba osiyanasiyana osiyana. Kulemba komwe kunandichititsa kuganiza pazinthu izi ndi Lazer et al. (2009) , Groves (2011) , Howison, Wiggins, and Crowston (2011) , boyd and Crawford (2012) , SJ Taylor (2013) , Mayer-Schönberger and Cukier (2013) , Golder and Macy (2014) , Ruths and Pfeffer (2014) , Tufekci (2014) , Sampson and Small (2015) , K. Lewis (2015b) , Lazer (2015) , Horton and Tambe (2015) , Japec et al. (2015) , ndi Goldstone and Lupyan (2016) .

M'mutu wonsewu, ndagwiritsa ntchito njira zamagetsi , zomwe ndikuganiza sizigwirizana nawo. Mawu ena otchuka omwe amawonekera pa digito ndi zojambulajambula (Golder and Macy 2014) , koma monga Hal Abelson, Ken Ledeen, ndi Harry Lewis (2008) akunena, mawu oyenerera kwambiri ndizojambula za digito . Mukamapanga mapazi, mumadziwa zomwe zikuchitika ndipo mapazi anu sangathe kuwonekera kwa inu enieni. Zomwezo sizowona pazochitika zanu zamagetsi. Ndipotu, mukusiya nthawi zonse zomwe muli ndi chidziwitso chochepa. Ndipo, ngakhale kuti zitsanzozi sizikhala ndi dzina lanu pa iwo, nthawi zambiri zimagwirizanitsidwa kwa inu. Mwa kuyankhula kwina, iwo ali ngati zolemba zala: osadziwika ndi kudziwika patokha.

• Big (gawo 2.3.1)

Kuti mudziwe zambiri chifukwa chake ma datasti akuluakulu amachititsa kuti mayesero apangidwe, onani M. Lin, Lucas, and Shmueli (2013) ndi McFarland and McFarland (2015) . Nkhanizi ziyenera kutsogolera ochita kafukufuku kuti azitha kuganizira mozama kusiyana ndi chiwerengero cha ziwerengero.

Kuti mudziwe zambiri zokhudza momwe Raj Chetty ndi anzake adzalandirira mauthenga a msonkho, onani Mervis (2014) .

Makampani akuluakulu angapangitsenso mavuto omwe amakumana nawo omwe sagwiritsidwe ntchito ndi kompyuta imodzi. Chifukwa chake, ochita kafukufuku omwe amapanga ma kompyuta akuluakulu nthawi zambiri amafalitsa ntchito pa makompyuta ambiri, zomwe nthawi zina zimatchedwa pulogalamu yofanana . Kuti muyambe kulumikiza mapulogalamu ofanana, makamaka chinenero chotchedwa Hadoop, onani Vo and Silvia (2016) .

• Nthawi zonse (gawo 2.3.2)

Poganizira nthawi zonse-pa deta, ndikofunika kulingalira ngati mukuyerekezera anthu omwewo pa nthawi kapena ngati mukuyerekezera kusintha kagulu ka anthu; onani chitsanzo, Diaz et al. (2016) .

• Zosagwiritsiridwa ntchito (gawo lachidule : osagwira ntchito? )

Buku lachikale pazitsulo zosasinthika ndi Webb et al. (1966) . Zitsanzo zomwe zili m'bukuli zisanafike zaka za digito, koma zikuwunikirabe. Kwa zitsanzo za anthu akusintha khalidwe lawo chifukwa cha kupezeka kwa masisitere, onani Penney (2016) ndi Brayne (2014) .

Reactivity ikugwirizana kwambiri ndi zomwe ochita kafukufuku amaitanitsa zotsatira (Orne 1962; Zizzo 2010) ndi zotsatira za Hawthorne (Adair 1984; Levitt and List 2011) .

• Zosakwanira (gawo 2.3.4)

Kuti mudziwe zambiri pazowonjezereka, onani Dunn (1946) ndi Fellegi and Sunter (1969) (mbiri) ndi Larsen and Winkler (2014) (amakono). Njira zofananazi zapangidwanso mu sayansi yamakompyuta pansi pa mayina monga data deduplication, chizindikiro chizindikiro, dzina kutanthauzira, zolemba kufotokoza, ndi zolembera mbiri kudziwika (Elmagarmid, Ipeirotis, and Verykios 2007) . Palinso njira zosungira zosungira zachinsinsi zomwe sizikufuna kufalitsa uthenga (Schnell 2013) . Facebook nayenso yakhazikitsa ndondomeko yolumikiza zolemba zawo ku khalidwe lovota; izi zinachitidwa kuti ndiyese kuyesa zomwe ndikukuuzani mu chaputala 4 (Bond et al. 2012; Jones et al. 2013) .

Kuti mudziwe zambiri pa zomangamanga, onani mutu 3 wa Shadish, Cook, and Campbell (2001) .

• Zosatheka (gawo 2.3.5)

Kuti mudziwe zambiri pa AOL search log debacle, onani Ohm (2010) . Ndimapereka malangizo othandizana ndi makampani ndi maboma mu chaputala 4 pamene ndikufotokoza zoyesera. Olemba ambiri adandaula za kafukufuku omwe akudalira deta Huberman (2012) , onani Huberman (2012) ndi boyd and Crawford (2012) .

Njira imodzi yabwino ofufuza yunivesite kupeza mwayi deta ndi ntchito pakampani ngati ntchitoyi kapena kuyendera kafukufuku. Kuwonjezera kuwapangitsa mwayi deta ndondomeko kumathandizanso kafukufuku kuti adziwe zambiri za momwe deta lisanalengedwe, chimene chiri chofunika kwa kusanthula.

Ponena za kupeza deta ya boma, Mervis (2014) akufotokoza momwe Raj Chetty ndi anzake adapezekanso mauthenga a msonkho omwe anagwiritsidwa ntchito mufukufuku wawo pankhani ya umoyo.

• Osayimira (gawo: osayimira)

Kuti mudziwe zochuluka pa mbiri ya "kuimirira" monga lingaliro, onani Kruskal and Mosteller (1979a) , Kruskal and Mosteller (1979b) , Kruskal and Mosteller (1979c) , ndi Kruskal and Mosteller (1980) .

Ndemanga zanga za ntchito ya Snow ndi ntchito ya Doll ndi Hill zinali zochepa. Kuti mudziwe zochuluka pa ntchito ya chisanu pa kolera, onani Freedman (1991) . Kuti mudziwe zambiri pa British Doctors Study onani Doll et al. (2004) ndi Keating (2014) .

Ofufuza ambiri adzadabwa kumva kuti ngakhale Doll ndi Hill atasonkhanitsa deta kuchokera kwa madokotala aakazi ndi madokotala osakwana zaka 35, iwo mwadala sanagwiritse ntchito deta iyi pakuyambanso koyamba. Pamene adakangana: "Popeza khansara yamapapu imakhala yosawerengeka pakati pa amayi ndi abambo oposa 35, owerengeka othandiza sangathe kupezeka m'magulu awa kwa zaka zingapo. Pa lipoti loyambirirali, tilingalira za amuna a zaka zapakati pa 35 ndi pamwamba. " Rothman, Gallacher, and Hatch (2013) , omwe ali ndi mutu wakuti" Chifukwa choyenera kuyanjana, "perekani ndemanga yowonjezereka ya kufunika kwa kupanga mwadala deta yosadziyimira.

Kusayimilira ndi vuto lalikulu kwa ofufuza ndi maboma omwe akufuna kufotokozera anthu onse. Izi sizikudetsa nkhaŵa makampani, omwe makamaka amagwiritsidwa ntchito kwa ogwiritsa ntchito. Kuti mudziwe zambiri za m'mene Statistics Netherlands Buelens et al. (2014) deta yaikulu, onani Buelens et al. (2014) .

Kwa zitsanzo za ofufuza omwe akudandaula za anthu omwe sali oimira chikhalidwe chachikulu, onani boyd and Crawford (2012) , K. Lewis (2015b) , ndi Hargittai (2015) .

Kuti mupeze kufotokoza mwatsatanetsatane kwa zolinga za kafukufuku ndi zofufuza za matenda, onani Keiding and Louis (2016) .

Kuti mudziwe zambiri pa kuyesa kugwiritsa ntchito Twitter kupanga zowonjezereka za ovota, makamaka nkhani ya chisankho cha Germany cha 2009, onani Jungherr (2013) ndi Jungherr (2015) . Pambuyo pa ntchito ya Tumasjan et al. (2010) ochita kafukufuku padziko lonse lapansi agwiritsira ntchito njira zowonongeka-monga kugwiritsira ntchito malingaliro kuti azitha kusiyanitsa pakati pazinthu zabwino ndi zolakwika za maphwando - kuti athandize deta ya Twitter kuti iwonetsere mitundu yambiri ya chisankho (Gayo-Avello 2013; Jungherr 2015, chap. 7.) . Momwe Huberty (2015) akufotokozera mwachidule zotsatira za zoyesayesazi kuti awonetse chisankho:

"Njira zonse zodziwiratu zokhudzana ndi chikhalidwe cha anthu zakhala zikulephera pamene zikutsatira zochitika zowonongeka. Zolakwitsa izi zimawonekera chifukwa cha zofunikira zamagulu a anthu, m'malo mwa zovuta kapena zovuta. Mwachidule, chitukuko chatsopano sichiti, ndipo mwinamwake sichidzatero, kupereka chithunzi chokhazikika, chosasamala, choyimira choyimira cha electorate; ndipo zitsanzo zabwino zokhudzana ndi chikhalidwe cha anthu sizikhala ndi chidziŵitso chokwanira chothetsera mavutowa. "

Mu chaputala 3, ndidzalongosola zitsanzo ndi kulingalira mwatsatanetsatane. Ngakhale ngati deta siyimilira, muzochitika zina, ikhoza kulemedwa kuti iwonetsekere bwino.

• Kuwongolera (gawo lachiyimire: kuthamanga? )

Kuthamanga kwadongosolo ndi kovuta kwambiri kuwona kuchokera kunja. Komabe, polojekiti ya MovieLens (yomwe ikufotokozedwa kwambiri mu chaputala 4) yakhala ikugwiritsidwa ntchito kwa zaka zoposa 15 ndi gulu lofufuza kafukufuku. Choncho, adatha kufotokozera ndi kugawana zambiri zokhudza njira yomwe dongosololi lasinthira m'kupita kwa nthawi komanso momwe izi zingakhudzira kusanthula (Harper and Konstan 2015) .

Akatswiri ambiri adayang'ana ku Twitter: Liu, Kliman-Silver, and Mislove (2014) ndi Tufekci (2014) .

Njira imodzi yogwiritsira ntchito chiwerengero cha anthu ndikutenga gulu la ogwiritsa ntchito, zomwe zimapangitsa ochita kafukufuku kuti aphunzire anthu omwewo mochedwa, onani Diaz et al. (2016) .

• Zosokonezeka mwamtendere (gawo ¿sec: mwangwiro-zosokonezeka? )

Ndinayamba kumva mawu akuti "algorithmically confounded" omwe adagwiritsidwa ntchito ndi Jon Kleinberg m'nkhani, koma mwatsoka sindinakumbukire kuti ndi liti pamene nkhaniyo inaperekedwa. Nthawi yoyamba imene ndinaona kuti inalembedwa inali ku Anderson et al. (2015) , yomwe ndi zokambirana zosangalatsa za momwe zida zogwiritsiridwa ntchito ndi malo ochezera amachepetsa mphamvu za akatswiri kuti azigwiritsa ntchito deta kuchokera pa webusaitiyi kuti aphunzire zosankha za anthu. Chisamaliro chimenechi chinayambitsidwa ndi K. Lewis (2015a) poyankha Anderson et al. (2014) .

Kuphatikiza pa Facebook, Twitter imalimbikitsa anthu kuti ogwiritsa ntchito azitsatira malingaliro a kutseka kwa triadic; onani Su, Sharma, and Goel (2016) . Kotero msinkhu wa kutseka kwa triadic pa Twitter ndi kuphatikizapo chizoloŵezi cha umunthu cha kutsekedwa kwa triadic ndi chizoloŵezi china chokonzekera cholimbikitsa kulimbitsa katatu.

Kuti mudziwe zambiri zokhudza chiwonetsero-makamaka chiganizo chakuti ziphunzitso zina za sayansi ndi "injini osati makamera" (mwachitsanzo, iwo amapanga dziko osati kungofotokoza) -kuwona Mackenzie (2008) .

• Dothi (gawo 2.3.9)

Mabungwe a ziwerengero za boma akuyitanitsa deta yoyeretsa mawerengedwe owonetsera deta . De Waal, Puts, and Daas (2014) amafotokoza njira zowonetsera deta zomwe zasankhidwa kuti azifufuza deta ndikuyang'ana momwe akugwiritsira ntchito pazinthu zazikulu, komanso Puts, Daas, and Waal (2015) amapereka malingaliro ofanana omvera ambiri.

Kuti mudziwe zambiri za social bots, onani Ferrara et al. (2016) . Kwa zitsanzo zina za maphunziro omwe adafuna kupeza spam mu Twitter, onani Clark et al. (2016) ndi Chu et al. (2012) . Pomaliza, Subrahmanian et al. (2016) afotokozera zotsatira za DARPA Twitter Bot Challenge, mgwirizano waukulu womwe umagwiritsidwa ntchito poyerekeza njira zoyang'anira bots pa Twitter.

• Zowonongeka (gawo 2.3.10)

Ohm (2015) ndemanga kafukufuku wakale pa lingaliro lodziwika bwino ndipo limapereka mayeso osiyanasiyana. Zinthu zinayi zomwe akukonzekera ndi kukula kwa kuvulaza, kuthekera koopsa, kupezeka kwachinsinsi, komanso ngati chiopsezochi chimaonetsa mavuto akuluakulu.

• Kuwerengera zinthu (gawo 2.4.1)

Kuphunzira kwa tekisi ku Farber ku New York kunayambira pa phunziro lakale la Camerer et al. (1997) yomwe idagwiritsa ntchito mapepala awiri oyendera mapepala osiyana. Kufufuza koyambirira kuja kunawona kuti oyendetsa galimoto ankawoneka kuti akuwombera opezapo: iwo ankagwira ntchito zochepa masiku omwe malipiro awo anali apamwamba.

Mu ntchito yotsatira, Mfumu ndi anzake adayang'ananso ku China (King, Pan, and Roberts 2014, [@king_how_2016] ) . Kuti mupeze njira yowunikira kuwunika pa intaneti ku China, onani Bamman, O'Connor, and Smith (2012) . Kuti mudziwe zambiri pazowerengera zomwe zikugwiritsidwa ntchito ku King, Pan, and Roberts (2013) kuti muyese kulingalira za ma 11 miliyoni, onani Hopkins and King (2010) . Kuti mudziwe zambiri pa maphunziro oyang'aniridwa, onani James et al. (2013) (zochepa) ndi Hastie, Tibshirani, and Friedman (2009) (owonjezera).

• Kuwonetseredwa ndi kufalitsa tsopano (gawo 2.4.2)

Chiwonetsero ndi gawo lalikulu la sayansi ya zamalonda (Mayer-Schönberger and Cukier 2013; Provost and Fawcett 2013) . Mtundu umodzi wa chidziwitso chomwe anthu ambiri amachititsa kuti anthu azidziŵira bwino, ndizowonetsa anthu; onani, mwachitsanzo, Raftery et al. (2012) .

Zotsatira za Google Flu sizinali zoyamba kugwiritsa ntchito deta yowunikira mpaka kufalikira kwa nthendayi yamakono. Ndipotu, ofufuza ku United States (Polgreen et al. 2008; Ginsberg et al. 2009) ndi Sweden (Hulth, Rydevik, and Linde 2009) apeza kuti mawu ena ofufuzira (mwachitsanzo, "chimfine") ananeneratu kuti dziko lonse lidzawonongeke deta isanatulutsidwe. Pambuyo pake, ntchito zina zambiri, zakhala zikuyesera kugwiritsa ntchito chiwerengero cha digito kuti chidziwitse matenda; onani Althouse et al. (2015) kuti mudziwe.

Kuphatikiza pa kugwiritsa ntchito deta yolongosola ndondomeko kuti muwonetsere zotsatira za thanzi, palinso ntchito yaikulu pogwiritsa ntchito data ya Twitter kuti liwonetsere zotsatira za chisankho; kwa maonekedwe awona Gayo-Avello (2011) , Gayo-Avello (2013) , Jungherr (2015) (chaputala 7), ndi Huberty (2015) . Tsopano kufotokoza kwa zizindikiro zachuma, monga zokolola zapakhomo (GDP), zimakhalanso zofala m'mabanki apakati, onani Bańbura et al. (2013) . tebulo 2.8 likuphatikizapo zitsanzo zochepa za maphunziro omwe amagwiritsa ntchito mtundu wina wa digito kulongosola mtundu wina wa chochitika padziko lapansi.

Pambuyo pake, Jon Kleinberg ndi anzake (2015) adanena kuti mavuto akuyambanso akugwera m'magulu awiri osiyana komanso kuti asayansi amayamba kuganizira za wina ndikumanyalanyaza. Tangoganizani wina wopanga mapulani, ndimutcha Anna, amene akulimbana ndi chilala ndipo ayenera kusankha ngati akulemba shaman kuti achite dansi la mvula kuti awonjezere mvula. Wopanga malamulo ena, ndidzamutcha Betty, ayenera kusankha ngati atenga ambulera kuti agwire ntchito kuti asamavutike panyumba. Ana ndi Betty akhoza kupanga chisankho chabwino ngati amvetsetsa nyengo, koma ayenera kudziwa zinthu zosiyana. Anna akuyenera kumvetsetsa ngati kuvina kwa mvula kumabweretsa mvula. Betty, kumbali inayo, safunikira kumvetsa kanthu kena kowopsa; iye akungofuna zolosera zolondola. Ochita kafukufuku wamagulu kawirikawiri amaganizira mavuto omwe Anna adakumana nawo-omwe Kleinberg ndi anzake amachitcha kuti "mavuto a ndondomeko ya mvula" monga chifukwa chokhala ndi mafunso. Mafunso ngati omwe Betty-omwe Kleinberg ndi anzake amagwiritsa ntchito amatcha "maambulera onga" mavutowo-angakhale ofunikira kwambiri, koma asamalimbikitse kwambiri ndi ochita kafukufuku.

• Zowonjezera kuyesa (gawo 2.4.3)

Magazini PS Political Science inali ndi nkhani yosiyirana yokhudza chidziwitso chachikulu, chidziwitso, ndi chiphunzitso, ndipo Clark and Golder (2015) mwachidule mndandanda uliwonse. Nyuzipepala ya Proceedings of the National Academy of Sciences ya ku United States of America inali ndi nkhani yosiyirana yonena za chiwerengero chachikulu ndi deta yaikulu, ndipo Shiffrin (2016) mwachidule mndandanda uliwonse. Kwa njira zophunzirira makina zomwe zimayesa kupeza zowonongeka zam'kati mkati mwazinthu zazikulu za deta, onani Jensen et al. (2008) , Sharma, Hofman, and Watts (2015) , ndi Sharma, Hofman, and Watts (2016) .

Malinga ndi zochitika zachilengedwe, Dunning (2012) amapereka chithandizo choyambirira, chotalikitsa buku ndi zitsanzo zambiri. Pofuna kukayikira zochitika zachilengedwe, onani Rosenzweig and Wolpin (2000) (Economics) kapena Sekhon and Titiunik (2012) (ndale). Deaton (2010) ndi Heckman and Urzúa (2010) amanena kuti kuyang'ana pa kuyesedwa kwachilengedwe kungachititse ochita kafukufuku kuganizira kuwonetsa zotsatira zosafunikira; Imbens (2010) akuwerengera mfundo izi ndi malingaliro abwino kwambiri a kufunika kwa kuyesedwa kwachirengedwe.

Pofotokozera momwe wofufuza angapitsidwire kuchokera kuwonetsa zotsatira za kukonzedwa ku zotsatira za kutumikira, ndinali kufotokoza njira yomwe imatanthawuza zinthu zosiyanasiyana . Imbens and Rubin (2015) , m'machaputala awo 23 ndi 24, amapereka ndemanga ndikugwiritsa ntchito loti lotengera chitsanzo. Zotsatira za ntchito ya usilikali pamakampani ena amatchulidwa nthawi zina amatchedwa caxal compact causal effect (CAcE) ndipo nthawi zina mankhwala am'derali amawopsa. Sovey and Green (2011) , Angrist and Krueger (2001) , ndi Bollen (2012) amapereka ndemanga zokhudzana ndi kugwiritsa ntchito zida zandale, zachuma, ndi zachikhalidwe, ndipo Sovey and Green (2011) amapereka "mndandanda wa owerenga" Kufufuza maphunziro pogwiritsa ntchito zosiyanasiyana.

Izi zikusonyeza kuti 1970 loti lotayira loti siinali yeniyeni; panali zochepa zapadera kuchokera ku zowona (Fienberg 1971) . Berinsky and Chatfield (2015) akunena kuti Berinsky and Chatfield (2015) kochepa kumeneku sikofunika kwenikweni ndipo kumakambirana kufunikira kochita mwakhama.

Malinga ndi zofanana, onani Stuart (2010) kuti muwone bwino, komanso Sekhon (2009) kuti muwone bwinobwino. Kuti mudziwe zambiri pofanana ndi mtundu wa kudulira, onani Ho et al. (2007) . Kupeza machesi amodzi kwa munthu aliyense nthawi zambiri kumakhala kovuta, ndipo izi zimayambitsa zovuta zambiri. Choyamba, pamene masewera enieni sakupezeka, ofufuza amafunika kusankha momwe angayezere mtunda pakati pa magulu awiri ndipo ngati mtunda wapadera uli pafupi kwambiri. Akatswiri amafuna kugwiritsa ntchito machesi angapo pa gulu lililonse lachipatala, chifukwa izi zingathe kuwonetseratu. Zonsezi, komanso ena, zifotokozedwa mwatsatanetsatane mu mutu 18 wa Imbens and Rubin (2015) . Onaninso Gawo II la ( ??? ) .

Onani Dehejia and Wahba (1999) kuti mudziwe njira zomwe zogwirizanitsa zimatha kupanga zowerengera zofanana ndi zomwe zakhala zikuyendetsedwa bwino. Koma, onani Arceneaux, Gerber, and Green (2006) ndi Arceneaux, Gerber, and Green (2010) kuti mupeze zitsanzo zomwe njira zofananako zinalephera kubwereza chizindikiro choyesa.

Rosenbaum (2015) ndi Hernán and Robins (2016) amapereka malangizo ena kuti azindikire kufananitsa kwabwino pakati pazidziwitso zazikulu.