ndemanga zina

Chigawo lakonzedwa kuti ntchito buku, osati muwerenge ngati mafotokozedwe.

  • Chiyambi (Ndime 2.1)

mtundu umodzi wa kusunga kuti sali m'gulu mu chaputala ichi ethnography. Zambiri pa ethnography mu malo digito kuona Boellstorff et al. (2012) , ndi zambiri pa ethnography mu wosanganiza malo digito ndi thupi kuona Lane (2016) .

  • Deta Big (Ndime 2.2)

Pamene inu repurposing deta, pali zidule awiri maganizo amene angakuthandizeni kumvetsa mavuto mwina inu kukumana. Choyamba, mungathe kuyesa kuyerekezera gulu lazidziwitso abwino vuto lanu ndi yerekezani kuti gulu lazidziwitso amene mumamwa. Kodi iwo ofanana ndi mmene zimasiyana bwanji? Ngati simunamukhululukire kusonkhanitsa deta yanu, pali ayenera kukhala kusiyana pakati pa zimene mufuna ndi zimene muli nazo. Koma, muyenera kusankha ngati kusiyana ndi yaing'ono kapena yaikulu.

Chachiwiri, kumbukirani kuti munthu analengedwa ndi kusonkhanitsa deta yanu pa zifukwa zina. Muyenera kuyesetsa kumvetsa maganizo awo. mtundu uwu n'zosiyana-zomangamanga kungatithandize kuzindikira mavuto ndi biases mu deta yanu repurposed.

Palibe single kugwirizana tanthauzo la "deta lalikulu", koma matanthauzo ambiri amaganizira za pa 3 motsutsana: (mwachitsanzo, buku, zosiyanasiyana, ndi mathamangidwe Japec et al. (2015) ). M'malo moika maganizo pa makhalidwe a deta, tanthauzo wanga kwambiri zimene bwanji deta lisanalengedwe.

kulolerana wanga deta boma utsogoleri mkati gulu deta lalikulu ndi pang'ono modabwitsa. Ena amene anapanga Pankhaniyi, monga Legewie (2015) , Connelly et al. (2016) , ndipo Einav and Levin (2014) . Zambiri za ubwino wa deta boma utsogoleri kufufuza, onani Card et al. (2010) , Taskforce (2012) , ndipo Grusky, Smeeding, and Snipp (2015) .

Kuti mukhale ndi maganizo a kafukufuku utsogoleri kuchokera mkati dongosolo boma zowerengera, makamaka Bureau Census US, onani Jarmin and O'Hara (2016) . Buku la kutalika mankhwala a kafukufuku utsogoleri mbiri pa Statistics Sweden, onani Wallgren and Wallgren (2007) .

Mu mutu, ine mwachidule poyerekeza kafukufuku miyambo monga General Social Survey (GSS) ndi chikhalidwe deta gwero media monga Twitter. Pakuti bwinobwino komanso kusamala poyerekeza pakati Kafukufuku chikhalidwe ndi deta chikhalidwe TV, onani Schober et al. (2016) .

  • Common makhalidwe a deta aakulu (Ndime 2.3)

Izi makhalidwe 10 deta yaikulu imene yatchedwa zosiyanasiyana njira zosiyana ndi zosiyanasiyana olemba osiyana. Kulemba kuti maganizo anga pa nkhani zimenezi ndi monga: Lazer et al. (2009) , Groves (2011) , Howison, Wiggins, and Crowston (2011) , boyd and Crawford (2012) , Taylor (2013) , Mayer-Schönberger and Cukier (2013) , Golder and Macy (2014) , Ruths and Pfeffer (2014) , Tufekci (2014) , Sampson and Small (2015) , Lewis (2015) , Lazer (2015) , Horton and Tambe (2015) , Japec et al. (2015) , ndipo Goldstone and Lupyan (2016) .

Mu chaputala ichi, ine ntchito mawu akuti kuda digito, yemwe ine ndikuganiza ndi ndale. Wina akuti ambiri tayamba digito ndi mapazi digito (Golder and Macy 2014) , koma Hal Abelson, Ken Ledeen, ndi Harry Lewis (2008) amanena, ndi yoyenera kwambiri akuti mwina digito Fingerprints. Pamene munakonza mapazi, mukudziwa zimene zikuchitika ndipo mapazi anu sangathe zambiri inachokera kwa inu. Yemweyo si zoona tayamba wanu digito. Ndipotu, akusiya kuda nthawi zonse zimene inu sadziwa kwambiri. Ndipo, ngakhale kuda izi alibe dzina lanu pa iwo, iwo nthawi zambiri zogwirizana kwa inu. M'mawu ena, iwo amakhala ngati Fingerprints: zosaoneka ndi panokha kudzizindikiritsa.

Big

Pakuti chifukwa chimene datasets lalikulu, amamasulira mayesero zowerengera mabvuto, onani Lin, Lucas, and Shmueli (2013) ndi McFarland and McFarland (2015) . nkhani zimenezi ayenera kutsogolera ofufuza kuganizira tanthauzo ochitachita osati tanthauzo zowerengera.

Nthawizonse pa

Mukamakambirana nthawi zonse, pa kafukufuku, n'kofunika kuganizira ngati inu kufanizitsa enieni anthu ofanana pa nthawi kapena ngati inu kufanizitsa gulu lina kusintha anthu; onani Mwachitsanzo, Diaz et al. (2016) .

Non-zotakasika

A buku tingachipeze powerenga pa miyeso si zotakasika ndi Webb et al. (1966) . The zitsanzo m'buku Pre-tsiku m'badwo digito, koma iwo adakali owala. Zitsanzo za anthu kusintha khalidwe lawo chifukwa choti misa anaziika, onani Penney (2016) ndi Brayne (2014) .

chosakwanira

Zambiri pa nkhani Mgwirizano, onani Dunn (1946) ndi Fellegi and Sunter (1969) (mbiri yakale) ndi Larsen and Winkler (2014) (masiku). Similar anapita akhala anayamba sayansi kompyuta pansi mayina monga deta deduplication, Mwachitsanzo chizindikiritso, dzina yofananira, kutsanzira kudziwika ndi kutsanzira mbiri kudziwika (Elmagarmid, Ipeirotis, and Verykios 2007) . Palinso chinsinsi chipulumutso likuyandikira kulemba Mgwirizano zomwe popanda kutenga panokha kudzizindikiritsa mudziwe (Schnell 2013) . Facebook nayenso anayamba chitani umakhalapo mayina awo khalidwe kuponya mavoti; ichi chidachitika kufufuza ndi kuyesera kuti ine ndikuuzani inu mu Chapter 4 (Bond et al. 2012; Jones et al. 2013) .

Zambiri pa kumanga n'loonadi, onani Shadish, Cook, and Campbell (2001) , Chapter 3.

kufikako

Zambiri pa AOL kufufuza chipika debacle, onani Ohm (2010) . I kupereka malangizo pa partnering ndi makampani ndi maboma mu Chapter 4 pamene ine pofotokoza zatsopano. Ambiri mwa olemba nkhawa za kafukufuku amene amadalira deta kufikako, 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.

Non-woimira

Non-representativeness ndi vuto lalikulu kwa akatswiri ndi maboma amene akufuna mawu za anthu onse. Izi ndi zochepa nkhawa makampani amene amagwiritsidwa lolunjika pa Intaneti awo. Kwambiri mmene Kafukufuku Netherlands asamalira nkhani ya si representativeness deta azamalonda, onani Buelens et al. (2014) .

Mu Chapter 3, ine pofotokoza zosankhidwazi ndi chiwerengero mwatsatanetsatane kwambiri. Ngakhale deta si a nthumwi, pa zinthu zina, iwo akhoza kulemetsedwa kubala ziwerengero wabwino.

kutengeka

System kulowerera ndi zovuta kwambiri kuona kuchokera kunja. Komabe, MovieLens ntchito (zomwe zinakambidwa zambiri Chapter 4) wakhala kuthamanga kwa zaka zoposa 15 ndi gulu kafukufuku akuphunzira. Choncho, iwo apeza ndi kugawana mfundo za momwe dongosolo wakhala kusanduka pa nthawi ndi mmene tingachitire zimakhudzira kusanthula (Harper and Konstan 2015) .

Ambiri mwa ophunzira akhala lolunjika pa kulowerera mu Twitter: Liu, Kliman-Silver, and Mislove (2014) ndi Tufekci (2014) .

Algorithmically manyazi

Ndinayamba kumva mawu akuti "algorithmically anadabwa" ntchito Jon Kleinberg mu nkhani. Mfundo yaikulu kumbuyo performativity kuti anthu ena sayansi mfundo chikhalidwe ndi "injini osati makamera" (Mackenzie 2008) . Ndiko kuti, iwo kwenikweni kuumba dziko osati kuulanda.

Akuda

Boma mabungwe zowerengera kuitana deta kuyeretsa, zowerengera deta kusintha. De Waal, Puts, and Daas (2014) pofotokoza zowerengera njira deta kusintha zithandize deta kafukufuku ndipo tione zimene mmene iwo applicable magwero aakulu deta, ndipo Puts, Daas, and Waal (2015) zinthu zina maganizo ofanana kwa omvera ambiri.

Pakuti zitsanzo za maphunziro lolunjika pa sipamu mu Twitter, Clark et al. (2016) ndi Chu et al. (2012) . Pomaliza, Subrahmanian et al. (2016) anafotokoza zotsatira za DARPA Twitter Bot Challenge.

tcheru

Ohm (2015) amakambirana kafukufuku kale pa lingaliro la zachinsinsi ndipo amapereka Mipikisano chinthu mayeso. The zinthu zinayi akumufunsira ndi: Mwina zoipa; Mwina zoipa; pamaso pa ubwenzi ndi chinsinsi; ndipo ngati chiopsezo kusonyeza nkhawa majoritarian.

  • Kuwerengera zinthu (Ndime 2.4.1)

Phunziro Farber la matakisi ku New York inachokera pa phunziro kale ndi Camerer et al. (1997) kuti ntchito zitatu zitsanzo osiyana mayiko pepala ulendo mapepala mapepala mitundu ntchito ndi madalaivala kulemba ulendo chiyambi nthawi, nthawi yotsiriza, ndipo likanatha. phunziro ili kale anapeza kuti madalaivala ankaoneka kuti chandamale earners: ankagwira ntchito kupatula pa masiku kumene malipiro awo apamwamba.

Kossinets and Watts (2009) anali pa chiyambi cha homophily Intaneti. Onani Wimmer and Lewis (2010) njira zosiyana vuto yemweyo amene amagwiritsa ntchito deta ku Facebook.

Ntchito wotsatira, Mfumu ndi anzake kuti zina kufufuza Intaneti kufufuza mu China (King, Pan, and Roberts 2014; King, Pan, and Roberts 2016) . Njira ina kuyeza Intaneti kufufuza mu China, onani Bamman, O'Connor, and Smith (2012) . Zambiri pa njira zowerengera monga munthu ntchito King, Pan, and Roberts (2013) chaka mawu a posts 11 miliyoni, onani Hopkins and King (2010) . Zambiri pa maphunziro ankaonetsetsa kuona James et al. (2013) (kupatula luso) ndi Hastie, Tibshirani, and Friedman (2009) (more luso).

  • Kuneneratu (Ndime 2.4.2)

Kulosera ndi gawo lalikulu la deta mafakitale sayansi (Mayer-Schönberger and Cukier 2013; Provost and Fawcett 2013) . Mtundu umodzi wa kulosera kuti anthu ambiri anachita mwa akatswiri chikhalidwe ndi zaziwerengero kulosera Mwachitsanzo Raftery et al. (2012) .

Google Matenda a Chimfine Trends sanali ntchito yoyamba kugwiritsa ntchito deta kusaka kwa nowcast kuchuluka fuluwenza. Ndipotu, ofufuza ku United States (Polgreen et al. 2008; Ginsberg et al. 2009) ndi Sweden (Hulth, Rydevik, and Linde 2009) apeza kuti mawu ena kufufuza (mwachitsanzo, "chimfine") ananeneratu national boma anaziika thanzi deta asanafike linatulutsidwa. Kenako zambiri, zinthu zina zambiri ankagwiritsa ntchito deta digito kufufuza kwa kudziwika anaziika matenda, onani Althouse et al. (2015) kuti ndemanga.

Kuwonjezera pa kugwiritsa ntchito deta digito kufufuza kulosera zakambidwa thanzi, apo wakhala pang'ono yaikulu ntchito data ndikuganizira Twitter kulosera zakambidwa chisankho; pakuti ndemanga kuona Gayo-Avello (2011) , Gayo-Avello (2013) , Jungherr (2015) (Ch. 7), ndipo Huberty (2015) .

Gwiritsani kusaka kwa akuneneratu kuchuluka fuluwenza ndi ntchito deta Twitter kulosera chisankho zitsanzo zonse za ntchito mtundu wa digito kufufuza kulosera mtundu wa chochitika mu dziko. Pali chiwerengero yaikulu kwambiri maphunziro kuti dongosolo lino ambiri. Table 2.5 zikuphatikizapo angapo zitsanzo zina.

Table 2.5: Tsankho mndandanda wa maphunziro ntchito kufufuza digito kulosera zinazake.
Intaneti kufufuza zotsatira tikalemba
Twitter Bokosi ndalama ofesi mafilimu mu US Asur and Huberman (2010)
Search zipika Sales mafilimu, nyimbo, mabuku ndi pakompyuta mu US Goel et al. (2010)
Twitter Dow Jones Industrial Avereji (US katundu msika) Bollen, Mao, and Zeng (2011)
  • Approximating zatsopano (Ndime 2.4.3)

Magazini PS Science Political anali yosiyirana pa kafukufuku wamkulu, causal kumasulira, ndi chiphunzitso ofunda, ndi Clark and Golder (2015) mwachidule aliyense zopereka. Magazini Zokambirana za Academy National wa Sciences la United States of America anali yosiyirana pa kumasulira causal ndi deta lalikulu, ndi Shiffrin (2016) mwachidule aliyense zopereka.

Kumbali ya zatsopano zachilengedwe, Dunning (2012) amapereka kwambiri buku kutalika mankhwala. Zambiri pa ntchito Vietnam kusodza mpikisano ngati kuyesera zachilengedwe, onani Berinsky and Chatfield (2015) . Pakuti njira makina kuphunzira kuti amafuna basi kupeza zatsopano zachilengedwe mkati mwa zinthu zazikulu deta, onani Jensen et al. (2008) ndi Sharma, Hofman, and Watts (2015) .

Kumbali ya yofananira, kwa review wosangalala, onani Stuart (2010) , ndi kwa review sizikuyenda bwino kuona Sekhon (2009) . Zambiri pa zofanana monga ngati kudulira, onani Ho et al. (2007) . Mabuku opereka chithandizo chabwino cha yofananira, onani Rosenbaum (2002) , Rosenbaum (2009) , Morgan and Winship (2014) , ndipo Imbens and Rubin (2015) .