2.5 Isiphelo
data Big yonke indawo, kodwa esebenzisa ngayo kunye nezinye iintlobo data lokuqwalasela uphando kwezentlalo kunzima. Ngokwamava am kukho into ngathi akukho mhlaba sasemini sikhululekile uphando: Ukuba anithanga wafakwa omninzi umsebenzi wokuqokelela data, ngoko ke, mhlawumbi uza kufuneka abekwe omninzi umsebenzi nokuhlalutya idata yakho okanye ucinga yintoni na umbuzo umdla nakucela data. Ngokusekelwe kwiimbono kwesi sahluko, ndicinga ukuba kukho iindlela ezintathu eziphambili ukuba imithombo yedatha enkulu iya kuba yeyona ixabisekile uphando kwezentlalo:
- empirically kugweba phakathi ezikhuphisanayo iingqikelelo theoretical. Imizekelo olu hlobo lomsebenzi ziquka Farber (2015) (abaqhubi New York Taxi) yaye King, Pan, and Roberts (2013) (kuhlolwa in China)
- eziphuculiweyo womlinganiselo zentlalo-nkqubo ngokuthi nowcasting. Umzekelo olu hlobo umsebenzi Ginsberg et al. (2009) (Google Flu ekwizinga).
- ukuqikelela iziphumo woko experiment zendalo ezingqameneyo. Imizekelo lo msebenzi. Mas and Moretti (2009) (peer iziphumo kwimveliso) kwaye Einav et al. (2015) (ifuthe ixabiso oqale nge fandesi eBay).
imibuzo emininzi ebalulekileyo kuphando zentlalo iza kuchazwa omnye kwaba bathathu. Nangona kunjalo, le ndlela ifuna ngokubanzi abaphandi ukuba kakhulu ukuba le datha. Yintoni eyenza Farber (2015) umdla izizathu theoretical ngenxa yokulinganisa. Le izizathu theoretical iphuma ngaphandle data. Ngenxa yoko, abo kakuhle ngokubuza kweentlobo ezithile zemibuzo yophando, imithombo yedatha enkulu zingaba nempumelelo kakhulu.
Okokugqibela, kunokuba ithiyori-eqhutywa kuphando reseach (leyo ibe ingqwalasela kwesi sahluko), sinako flip elishicilelwe nokudala ngcamango-oluqhutywa empirically. Oko kukuthi, ngalo beqokelela ngenyameko izibakala lokufumanisa ngophando, iipateni, neephazili, siya kwakha iingcamango ezintsha.
Le ndlela, indlela data-yokuqala ingcamango asinto intsha, laye iimbono ekulinywe Glaser and Strauss (1967) nekhwelo yabo ithiyori nisekelwe. Le ndlela data-yokuqala, nangona kunjalo, oko akuthethi "isiphelo ithiyori," njengoko sele ibangwe ezininzi nobuntatheli ngeenxa uphando kweminyaka digital (Anderson 2008) . Kunoko, njengoko bume data utshintsho, kufuneka silindele kwakhona balancing kobudlelwane phakathi kwengcingane data. Kweli hlabathi apho ukuqokelelwa kwedatha eninzi, iyavakala ukuqokelela kuphela data iingcamango bacebisa kuya kuba luncedo kakhulu. Kodwa, kwihlabathi apho izixa emikhulu data zifumaneka simahla kakade, iyavakala ukuba uzame indlela data-lokuqala (Goldberg 2015) .
Njengoko kubonisiwe kwesi sahluko, abaphandi ungafunda okuninzi ngokubukela abantu. Kwizahluko emithathu ezayo, ndiza ukuchaza indlela sifunda izinto ezininzi ezahlukileyo ukuba Singayenza ukuqokelelwa yethu data kwaye banxibelelane nabantu abaninzi ngqo ngokubuza imibuzo (Isahluko 3), imifuniselo ibaleka (Isahluko 4), kunye ngokuzibandakanya kwinkqubo yophando ngqo (Isahluko 5).
