Fungidzira Ndiani? 5 mienzaniso nei kubvisa mazita haisi sarudzo

fungidzira kuti ndiani mutambo

Sumo kuna Guess Ndiani

Fembera Ndiani? Kunyangwe ini ndine chokwadi chekuti vazhinji venyu munoziva mutambo uyu kubva kumashure mumazuva, heino kupfupisa kudzokorora. Chinangwa chemutambo: tsvaga zita remufananidzo wekatuni wakasarudzwa nemuvengi wako nekubvunza 'hongu' uye 'kwete' mibvunzo, senge 'munhu wacho anopfeka ngowani here?' kana 'munhu wacho anopfeka magirazi'? Vatambi vanobvisa vavhoterwa zvichibva pamhinduro yemupikisi uye vodzidza hunhu hunoenderana neyakavanzika hunhu hwemupikisi wavo. Mutambi wekutanga uyo anoverengera mumwe mutambi chakavanzika hunhu anokunda mutambo.

Iwe unazvo. Mumwe anofanirwa kuona munhu wacho kunze kwedhatabhesi nekungove nekuwana chete kune zvinoenderana hunhu. Muchokwadi, isu tinogara tichiona ichi chirevo cheGuess Ndiani akashandisa mukuita, asi ndokuzoshandirwa pamadhata akaumbwa nemitsara nemakoramu ane hunhu hwevanhu chaivo. Musiyano mukuru paunenge uchishanda nedata ndewekuti vanhu vanowanzo kushora kurerukirwa uko vanhu chaivo vanogona kuburitswa nekungova nehunhu chete.

Seizvo Guess Ndiani mutambo unoratidzira, mumwe munhu anogona kuziva vanhu nekuve nekugona kune mashoma chete hunhu. Iyo inoshanda semuenzaniso uri nyore wekuti nei uchibvisa chete 'mazita' (kana zvimwe zvinonongedza zvakanangana) kubva kune dhata rako rinokundikana senge nzira yekuzivisa. Mune ino blog, isu tinopa ina anoshanda kesi kukuzivisa iwe nezve zvakavanzika njodzi zvine chekuita nekubviswa kwemakoramu senzira yedata kusazivikanwa.

2) Kurwiswa kwekubatanidza: dhatabhesi rako rakabatanidzwa kune mamwe (ruzhinji) masosi edata

Njodzi yekurwiswa kwekubatana ndicho chikonzero chakakoshesa nei kungobvisa mazita kusingashande (zvekare) senzira yekuzivisa. Nekurwiswa kwekubatana, uyo anorwisa anosanganisa iyo data yekutanga nemamwe masosi anowanikwa kuitira kuti anyatsoziva mumwe uye adzidze (kazhinji kazhinji) ruzivo nezve uyu munhu.

Chakakosha apa ndiko kuwanikwa kweimwe dhata zviwanikwa zviripo izvozvi, kana zvinogona kuve zviripo mune ramangwana. Funga nezvako. Yakawanda sei yako data rako pachako inogona kuwanikwa paFacebook, Instagram kana LinkedIn iyo inogona kuve inogona kushungurudzwa pakurwadziswa kwekubatana?

Mumazuva apfuura, kuwanikwa kwedata kwaive kwakanyanya kushomeka, izvo zvinotsanangura kuti nei kubviswa kwemazita kwaikwana kuchengetedza zvakavanzika zvevanhu. Dhata shoma iripo inoreva mikana shoma yekubatanidza data. Nekudaro, isu tave ikozvino (vanoshingairira) vatori vechikamu muhupfumi hunotungamirwa nedatha, uko huwandu hwedata huri kukura padanho rekujekesa. Mamwe data, uye kugadzirisa tekinoroji yekuunganidza dhata kunotungamira kukuwedzera mukana wekubatana kwekurwiswa. Chii chaizonyorwa nemumwe mumakore gumi nezve njodzi yekurwiswa kwekubatana?

Mufananidzo 1

Exponentially kukura dhata ichokwadi

Kuwanda kwedata

Nyaya yekudzidza

Sweeney (2002) akaratidzira mupepa redzidzo kuti akakwanisa sei kuona nekutora data rezvehutano kubva kune vanhu zvichibva pakubatanidza ruzhinji rwe data 'rwekushanyira chipatara' kune anowanikwa pachena anyoresa kuvhota muUnited States. Ose madhateseti pane ayo anofungidzirwa kuve anonyatso kuzivikanwa kuburikidza nekubviswa kwemazita uye zvimwe zvinonongedza zviri pachena.

Mufananidzo 2

Linkage kurwisa mukuita

Kubatanidza Kurwisa

Zvichienderana nezvikamu zvitatu chete (1) Zip Code, (2) Gender uye (3) Zuva rekuzvarwa, akaratidza kuti 87% yevagari vemuUS vanogona kuzivikanwa zvakare nekufananidza zvataurwa pamusoro apa kubva kumaseteti maviri. Sweeney akabva adzokorora basa rake aine 'nyika' seimwe nzira ye 'Zip Code'. Pamusoro pezvo, akaratidzira kuti 18% yehuwandu hwevanhu vemuUS vanokwanisa kuzivikanwa chete nekuwana mukana kune dhata rine ruzivo nezve (1) nyika yekumusha, (2) mukadzi uye (3) zuva rekuzvarwa. Funga nezve zvataurwa pamusoro izvi sosi, senge Facebook, LinkedIn kana Instagram. Nyika yako, murume kana murume uye zuva rekuzvarwa rinoonekwa, kana vamwe vashandisi vanokwanisa kuitora?

Mufananidzo 3

Mhedzisiro yaSweeney

Quasi-zvitupa

% yakasarudzika yakasarudzika yeUS huwandu hwevanhu (248 mamirioni)

5-manhamba ZIP, jenda, zuva rekuzvarwa

87%

nzvimbo, jenda, zuva rekuzvarwa

53%

nyika, jenda, zuva rekuzvarwa

18%

Uyu muenzaniso unoratidza kuti zvinogona kuve zvinoshamisa zvinoshamisa kuti de-kusazivisa vanhu mune inoita kunge isingazivikanwe data. Kutanga, chidzidzo ichi chinoratidza hukuru hukuru hwenjodzi, se 87% yehuwandu hweUS inogona kuzivikanwa nyore kushandisa mashoma maitiro. Chechipiri, iyo yakajeka yekurapa data muchidzidzo ichi yaive yakanyanya kunetseka. Mienzaniso yeakafumurwa data revanhu kubva kuchipatara chekushanyira dheti inosanganisira dzinza, kuongororwa uye mishonga. Hunhu hunogona kungochengetwa hwakavanzika, semuenzaniso, kubva kumakambani einishuwarenzi.

3) Vanhu vane ruzivo

Imwe njodzi yekubvisa zvinonongedza chete, senge mazita, anomuka kana vanhu vane ruzivo vaine ruzivo rwepamusoro kana ruzivo nezve hunhu kana hunhu hwevamwe vanhu vari mudhatabhesi.. Zvichienderana neruzivo rwavo, anorwisa anogona kuzokwanisa kubatanidza mamwe marekodhi edhata kune vanhu chaivo.

Nyaya yekudzidza

Muenzaniso wekurwiswa kwedhataset uchishandisa ruzivo rwepamusoro inyaya yeNew York tekisi, uko Atockar (2014) akakwanisa kuburitsa vanhu chaivo. Iyo yakashandiswa dhataset yaive neyese nzendo dzetaxi muNew York, yakafumiswa nezvakakosha hunhu senge kutanga mabhaisikopu, magumo ekubatanidza, mutengo uye tip of the ride.

Mumwe munhu ane ruzivo anoziva New York akakwanisa kuwana nzendo dzetaxi kukirabhu yevakuru 'Hustler'. Ne kusefa iyo 'yekupedzisira nzvimbo', akaisa chaiyo chaiyo kero kero uye nekudaro akaziva akasiyana siyana anowanzo shanya. Saizvozvo, mumwe anogona kutora kukwira tekisi kana kero yekumba yemunhu yaizivikanwa. Iyo nguva nenzvimbo yeanoverengeka anozivikanwa emamuvhi nyeredzi akawanikwa pane makuhwa nzvimbo. Mushure mekubatanidza iyi ruzivo kune iyo NYC tekisi data, zvaive nyore kuwana avo vatasvi vematekisi, iyo mari yavakabhadhara, uye kuti vanga vabhadharisa here.

Mufananidzo 4

Munhu anoziva

kudonhedza kunongedzera Hustler

Bradley Cooper

taxi uye mepu

Jessica Alba

kutevedza mamepu

4) Dhata sechigunwe chemunwe

Mutsara wakajairika wekupokana ndeyekuti 'iyi data haina maturo' kana 'hapana-munhu anogona kuita chero chinhu neiyi data'. Uku kunowanzo kuve kufunga kusiri iko. Kunyangwe iyo isina mhosva data inogona kuumba yakasarudzika 'fingapurindi' uye ishandiswe kuzivisa zvekare vanhu. Ndiyo njodzi inotorwa kubva mukutenda kuti iyo data pachayo haina maturo, asi zvisiri.

Ngozi yekuzivikanwa ichawedzera pamwe nekuwedzera kwedata, AI, uye zvimwe zvishandiso nemaalgorithms ayo anoita kuti kufumurwa kwehukama hwakaoma mu data. Nekudaro, kunyangwe kana dhatabhesi rako risingakwanise kuvhurwa izvozvi, uye zvingangodaro zvisina basa kuvanhu vasina mvumo nhasi, zvinogona kunge zvisiri mangwana.

Nyaya yekudzidza

Muenzaniso wakanaka inyaya iyo iyo Netflix yaida kuunganidza dhipatimendi reR & D nekuunza yakavhurika makwikwi eNetflix yekuvandudza yavo yekurumbidza bhaisikopo system. 'Iyo inovandudza kubatana kwekusefa algorithm kufanotaura mushandisi ratidziro yemafirimu inohwina mubairo weUS $ 1,000,000'. Kuti utsigire vanhu vazhinji, Netflix yakaburitsa dhatasheti ine chete anotevera hunhu hunokosha: mushandisiID, bhaisikopo, zuva regiredhi uye giredhi (saka hapana rumwe ruzivo pamusoro pemushandisi kana firimu pachayo).

Mufananidzo 5

Dataset chimiro Netflix mutengo

UserID Bhaisikopo Zuva regiredhi giredha
123456789 Basa hazviiti 10-12-2008 4

Mukuzvimiririra, iyo data yakaratidza kunge isina maturo. Kana uchibvunza mubvunzo kuti 'Kune chero ruzivo rwevatengi mune dhata rinofanirwa kuchengetwa rakavanzika?', Mhinduro yaive:

 'Kwete, ruzivo rwese rwekuzivisa vatengi rwabviswa; zvese zvasara zviyero nemadheti. Izvi zvinotevera gwara redu rekuvanzika… '

Nekudaro, Narayanan (2008) anobva kuYunivhesiti yeTexas kuAustin akaratidza zvimwe. Iko kusanganiswa kwemakero, zuva regiredhi uye bhaisikopo remumwe munhu rinogadzira yakasarudzika bhaisikopo-chigunwe. Funga nezve yako Netflix maitiro. Iwe unofunga kuti vanhu vangani vakatarisa imwechete seti yemamuvhi? Vangani vakatarisa imwechete seti yemamuvhi panguva imwe chete?

Mubvunzo mukuru, sei kuenzanisa ichi chigunwe chemunwe? Zvaive zviri nyore. Zvichienderana neruzivo kubva kune inozivikanwa-yemafirimu-rating webhusaiti IMDb (Internet Movie Database), yakafanana fingapurindi inogona kuumbwa. Nekudaro, vanhu vanogona kuzivikanwa zvakare.

Nepo maitiro ekuona-bhaisikopo angave asingafungidzirwe semashoko akaomarara, funga nezvehunhu hwako - ungafunga here kana zvikaenda kuruzhinji? Mienzaniso yakapihwa naNarayanan mupepa rake ndezvematongerwo enyika (ratings pa 'Jesu weNazareta' uye 'Evhangeri yaJohani') uye zvido zvebonde (ratings pa 'Bent' uye 'Queer sevanhu') izvo zvinogona kuraswa nyore.

5) General Dziviriro Dhata (GDPR)

GDPR inogona kunge isingaite-inonakidza, kana bara resirivha pakati penyaya dzeblog. Asi, zvinobatsira kuti uwane tsananguro dzakananga kana uchigadzirisa data rako pachako. Sezvo iri blog riri pamusoro peyakajairika pfungwa isiriyo yekubvisa makoramu senzira yekuzivisa data uye kukudzidzisa iwe se data processor, ngatitangei nekutsvaga tsananguro yekusazivikanwa maererano neGDPR. 

Zvinoenderana neye recital 26 kubva kuGDPR, ruzivo rwusingazivikanwe runotsanangurwa se:

'ruzivo rwusinei neanozivikanwa kana kuzivikanwa munhu chaiye kana dhata remunhu rakashumirwa kusazivikanwa nenzira yekuti iyo data iri pasi haichagone kuzivikanwa.'

Sezvo imwe ichigadzirisa dhata yega ine chekuita nemunhu chaiye, chikamu chechipiri chete chedudziro chakakosha. Kuti uite zvinoenderana netsananguro, munhu anofanirwa kuona kuti iyo data subject (yega) haisi kana kusazivikanwa. Sezvinoratidzwa mune ino blog, zvisinei, zviri nyore zvinoshamisa kuti uzive vanhu zvichibva pane zvishoma hunhu. Saka, kubvisa mazita mudhatabhesi hakuenderane nedudziro yeGDPR yekusazivikanwa.

In mhedziso

Isu takadenha imwe inowanzo fungidzirwa uye, zvinosuruvarisa, ichiri kungoshandiswa nzira yedata kusazivikanwa: kubvisa mazita. Mu Guess Ndiani mutambo uye mimwe mimwe mienzaniso nezve:

  • Linkage kurwisa
  • Vanhu vane ruzivo
  • Dhata sechigunwe chemunwe
  • General Data Protection Regulation (GDPR)

zvakaratidzwa kuti kubvisa mazita kunokundikana sekusazivikanwa. Kunyangwe iyo mienzaniso iri inokatyamadza kesi, imwe neimwe inoratidza iko kuve nyore kwekuzvidzoreredza uye kukanganisa kunogona kukanganisa kuvanzika kwevanhu.

Mukupedzisa, kubviswa kwemazita kubva kune yako dhatabhesi hakukonzere data risingazivikanwe. Nekudaro, isu zvirinani tidzivise kushandisa ese ari maviri mazwi kuchinjana. Ndinovimba nemoyo wese kuti hauzoshandise nzira iyi yekusazivikanwa. Uye, kana uchiri kudaro, ita shuwa kuti iwe nechikwata chako munonzwisisa zvizere njodzi dzekuvanzika, uye munobvumidzwa kugamuchira njodzi idzi pachinzvimbo chevanhu vakanganiswa.

boka revanhu vachinyemwerera

Dhata ndeyekugadzira, asi timu yedu ndeyechokwadi!

Bata Syntho uye imwe yenyanzvi dzedu ichasangana newe nekumhanya kwechiedza kuti uongorore kukosha kweiyo synthetic data!

  • D. Reinsel, J. Gantz, John Rydning. Iyo Digitization yeNyika Kubva paEdge kusvika kuCore, Dhata Age 2025, 2018
  • L. Sweeney. k-kusazivikanwa: muenzaniso wekuchengetedza zvakavanzika. International Journal pamusoro pekusava nechokwadi, Fuzziness uye Knowledge-based Systems, 10 (5), 2002: 557-570
  • L. Sweeney. Zvirinyore huwandu hwevanhu Kazhinji Kuzivisa Vanhu Zvakasarudzika. Carnegie Mellon University, Data Yakavanzika Kushanda Pepa 3. Pittsburgh 2000
  • P. Samarati. Kuchengetedza Vapinduri 'Zviziviso muMicrodata Kuburitswa. IEEE Shanduko paRuzivo neData Injiniya, 13 (6), 2001: 1010-1027
  • Atockar. Kutasva neNyeredzi: Yakavanzika Yevatakuri muNYC Taxicab Dataset, 2014
  • Narayanan, A., & Shmatikov, V. (2008). Robust de-kusazivisa emahombe madhata mashoma. Mukuenderera mberi - 2008 IEEE Symposium pamusoro peChengetedzo uye Kuvanzika, SP (map. 111-125)
  • General Dziviriro Dziviriro