ICandelo leDatha. unyaka ka-2013. Ukubuyisela emva

Ngonyaka we-2013 IBS, eyayibonakala ngathi iyadala ICandelo leDatha, wandicela ukuba ndenze i-braindump enjalo (ngokukodwa ngokusekelwe kumava entsebenziswano kunye nabathengi beoyile kunye negesi) malunga nommandla wengxaki yeDatha enkulu, kunye neDatha ngokubanzi. Ndiye ndayibona emva kweminyaka eyi-7 ndacinga ukuba iyahlekisa. Ezinye izinto zicacile. Abanye bajika bangabi yinyani ngokupheleleyo, kodwa ... iminyaka eyi-7 idlulile.

Ndabhala ngesiNgesi kwaye ngoku ndacinga ukuyiguqulela kwisiRashiya. Kuthekani ukuba kukho into esabalulekile ngoku? (Ndiza kuguqulela iibhulethi, kodwa ushiye iimpawu ngesiNgesi ngenxa yobuvila. Uhlaza lulungile, olubomvu luyingozi, blue liphupha).

Ndiza kuwenza ngokusesikweni amagqabantshintshi ukusuka β€œnamhlanje” IsiTaliyaniukuze icace kwaye icace.

Ngoko, IDATHA! Idata yethu...

ICandelo leDatha liCandelo leGazi, kuba idatha inokuthelekiswa, umzekelo, kwigazi elihamba ngemithambo kunye nemithambo yegazi yomzimba woshishino. Nangona kunjalo, nangona igazi lifana, izinto eziphilayo zihluke kwaye ngoko ke ukwenziwa kwemveliso kunzima kakhulu, kodwa ikwamele ithuba lophuhliso.

Kukho abantu abathi idatha itsibela ngqo emehlweni abo - aba Siyiyo.
Kwaye kukho abantu abathi, ngelishwa, abayiboni indawo yedatha ingenanto. Oku, kwakhona, yeha, yeyethu Abathengi!

ICandelo leDatha. unyaka ka-2013. Ukubuyisela emva

Kwaye ke, imigaqo yeshishini...

  1. Siyathengisa ishishini, kwaye akunjalo ИВ (Ngamana zonke iingcali ze-IT zingandixolela kanye) kuba sisombulula iingxaki zehlabathi, kwaye, imali eninzi.
  2. Zonke iingxaki zeshishini zigxile kwimixholo yoshishino oluthe ngqo kwaye ziya kufuna okwaneleyo iingcali.
  3. Iinzame zokungqina ixabiso le "data" okanye, kunzima ngakumbi, ixabiso "lolawulo lwedatha" kwishishini kukubandezeleka okungunaphakade kunye neentlungu. Ngokusisiseko, kufana nokuza emntwini oziva kamnandi uze uthi: β€œMfondini, siza kunyanga igazi lakho ngoku, yaye, mfondini, liyabiza!”
  4. "Iphupha lam elimanzi" kukuthengisa "ukukhutshwa kwedatha" kunye "nohlalutyo" ngaphakathi kwemodeli ye-SaaS amashishini amancinci naphakathingubani oye wenyuka kwiinkonzo zefu ze-123 kunye ne-interfaces epholileyo: ulawulo lweprojekthi, i-helpdesk, i-accounting, i-CRM, i-payroll, ingxelo yexesha, ukuthengisa, ... uyibiza ngegama, kwaye wazingcwaba kwidatha. Youcalc kunye nabaphumeleleyo (mhlawumbi azisekho) Oku Kulungile!
  5. Khangela abantu abathanda ukunkcunkca "qhwaba" ngedatha. Zinqabile kwaye zingaqhelekanga (njengamagqabi eti), kodwa zingundoqo kwishishini. Imbongi, umzekelo, inokuba kakuhle kakhulu ekunxibelelaneni.
  6. Iinjineli efunekayo! Kufuneka ukuguqula iingxaki ezitsalwa yiCrunchers kwidatha zibe zizisombululo. Kwaye impumelelo okanye ukungaphumeleli kwesigqibo kuxhomekeke ngokupheleleyo kubo.
  7. Phuhliso Vula Umnikezi Iiprojekthi zinexabiso elikhulu kwaye zenza kube lula "ukudibanisa" izisombululo ezinzima ukusuka ekuqaleni.
  8. Kodwa ... akufanele sikhohlwe ukuba i-Hadoop yilayibrari, kwaye i-Lucene nayo ilayibrari, kunye nomgama ophakathi ilayibrari kunye nemveliso yoshishino kakhulu!
  9. Izisombululo ezakhiweyo ziya kufuneka zilungelelaniswe ngokuphawulekayo, kuba ukumodareyitha ΠΈ imfezeko - amanqaku aphambili.
  10. Agile (UThixo andixolele) bubuchule obuphambili ekusebenzisaneni nomthengi kunye nokuqinisekisa iingcamango, apho kuya kubakho abaninzi.
  11. Kuyenzeka ngakumbi kwaye kuyimfuneko ukukhupha zonke iikhowudi kunye ne-UI. Lonke uhlalutyo lweshishini kunye neenkcukacha ngasemva kufuneka uhambe ngaphakathi kwaye ithathwa njengesakhono esingundoqo.
  12. Abenzi bezigqibo zoshishino kufuneka bahlale "benolwazi" malunga imfuneko yokusebenza ngokufanelekileyo ngedatha kwaye rhoqo ukukhangela iindlela ezintsha zokuzihlalutya. Ukudityaniswa kwezakhono zobugcisa kunye nezoshishino zabasebenzi bethu kuya kunceda ukuphakamisa umgangatho wombutho uphela.
  13. Internet - kukho umthombo ongapheliyo wempembelelo (zazingekho iikati ezingako ngelo xesha) ngokunxulumene neendlela zolawulo lwedatha yeshishini, nangona iinjongo kunye nobubanzi bohluka kakhulu.

ICandelo leDatha. unyaka ka-2013. Ukubuyisela emva

Iimpawu zetekhnoloji ...

  1. Kukho amandla amakhulu ophuhliso kwi ukwenza lula indlela idatha eboniswa ngayo ebantwini. Ungabiza oku ngegama elithi "iPhonization".
  2. Ngaphandle kwento yokuba abathengisi be-BI bathi ngokuthe ngqo zisa uhlalutyo kubasebenzisi bokugqibela, (kwaye ngokuqinisekileyo bayahamba kweli cala) - ukuphumelela akukenzeki. Abantu abaqondi kakuhle multidimensional idatha.
  3. Ujongano lomsebenzisi olumele ngaphezulu okanye ngaphantsi kobunzima, idatha eyakhiwe ngokukhululekileyo ngaphakathi ubuso ngobuso ifomu - ikwabonisa inani elingapheliyo leengxaki. Isiphelo: i-flatter ingcono.
  4. Iqonga elakhiwe ngesiseko sokutsalwa kwedatha ngokuzenzekelayo kwimithombo (engasoloko yenzelwe ukutsalwa okunjalo) ixhomekeke kakhulu kwimithombo, ukuzinza kweziqhagamshelo, kunye neziseko. Iqonga (umthunywa) liya kuhlala lityholwa ngokusilela ukuhambisa iziphumo. Ithemba – inkunzi yolu hlobo lwamaqonga. Inkunzi ekunzima ukuyizuza kwaye kulula ukuyilahleka.
  5. Ukusuka kumbono woshishino, akukho mahluko phakathi kohlalutyo lweDatha enkulu kunye Idatha nje. Rhoqo emva kwamanani alula njengo 2x2 kukho izigidi zeedola zamathuba. Umzekelo omhle yidatha ekupheleni kobomi bezinto ezisisiseko kwishelufu yaseNorway. Inini yonke imihla yee-caps ezizayo. Ukulungiswa kwazo zonke izixhobo kwafakwa kwiaxis enye kwaye bafumanisa ukuba kwiminyaka eyi-N ishelufu yeArmagedon yayiza kufika - enye indoda esisityebi yaphakama esitulweni sayo yaza yaqubuda ngaphandle kwegumbi inamazwi athi: β€œUxolo, andikwenzi. Ndinexesha elininzi, kufuneka ndilungiselele inqanawa. ” …
  6. I-Excel, kwaye ngokusisiseko ingcaciso yetheyibhile ecacileyo nemfutshane, inamandla amakhulu kunye nekamva elikhulu. Ndiyakholelwa kwiitafile ezintle (kwaye usenjenjalo) kwaye yiloo nto!
  7. Isaphetha esiphambili sayo yonke le "analytics" yi Isigqibo esizenzekelayo. Kukho amathuba amakhulu, kodwa kunye nemingcipheko ephezulu, yiyo loo nto amathuba atyebileyo, yingakho kukho imingcipheko, yingakho kukho amathuba, yingakho i-toffee ... πŸ™‚ Ukulawulwa kakuhle kwe-drilling, umzekelo ...
  8. Ukuba "ukudityaniswa" luphawu oluphambili, ngoko ke idatha kufuneka iboniswe njengenkonzo. ABANYE imithetho, kodwa akufuneki silibale malunga nokwenza ngcono imveliso, ngoku ehlala idini ngenxa yokudityaniswa njengoko amandla ekhompyuter eqhubeka ekhula.
  9. Idatha enkulu - oku kufuneka kubekwe kwindawo, kutsalwe, kubekwe emgangathweni phambi kokujongana nayo nayiphi na imiba yeshishini. Idatha enkulu incinci, kodwa iingxaki ngayo zinkulu! Njengoko abazalwana be-semantics besitsho, i-50% yeengxaki zehlabathi zibangelwa kukuba abantu babiza izinto ezifanayo ngamagama ahlukeneyo, kwaye enye i-50% kukuba babiza izinto ezahlukeneyo ngegama elinye.
  10. Naluphina i-encapsulation kwinqanaba lokugcina, linciphisa ukuvuleka kwesisombululo kwaye likhokelela kwi-SILO-fication. Kulungile ukuba ungumthengisi omkhulu, kungenjalo kunjalo-njalo. (Apha sithetha, ngokuqinisekileyo, kungekhona malunga nenqanaba lebhloko kwaye kungekhona nge-AWS S3, eyayisele ineminyaka eyi-6 ubudala ngoko, kodwa malunga neefayile.).
  11. Imodeli yobudlelwane data ayisekho umhlobo wethu. I-RDF kunye nexabiso elingundoqo-lipholile! Siye sabona ukuguqulwa komlingo koovimba beenkcukacha zobudlelwane kunye neemodeli zeetafile ze-2000 kwiitafile ze-15, kwaye akukho namnye wabasebenzisi abalahlekelwe nantoni na.
  12. I-Intanethi isebenza ngenxa yokuba ikhona URL njengendlela emanyeneyo yokujongana. Ukubaluleka kwe-URL okanye kunoko I-URI kuba izibonelelo zolwazi lweshishini kunzima ukuba overestimate.
  13. Iimayini zombhalo kunye ne-NLP zidumile. Kwi-Intanethi. Kodwa nakwicandelo leenkampani, impumelelo enkulu inokufezekiswa ngokukhupha idatha ecwangcisiweyo kwidatha yenkampani engacwangciswanga.
  14. Ulungelelwaniso phakathi kwedatha ecwangcisiweyo kunye nolwazi olukhutshwe kwidatha engacwangciswanga, i.e. iifayile - uhlalutyo Klondike.
  15. Xa ukhupha idatha, ungalibali malunga namalungelo kunye amalungelo okushicilela.
  16. Inkampani yokutsalwa kwedatha kufuneka yenze aisebe labahlaseli, ngengqiqo efanelekileyo yegama. Ikhuthazwe ngumlo onyukayo ngokuchasene neenkqubo zokhuseleko zebhot yeYellow Pages crawler.
  17. Ngaphambi kokuba usebenze ngedatha, kuyimfuneko ukuba "yabona" ngokupheleleyo. Kunzima ukucacisa. Iifom zetheyibhile zifika engqondweni. Kwabanye, ukubonakaliswa kwemizobo, kodwa nayiphi na igrafu sele iyitoliko. Enye indlela okanye enye ... "bona"!
  18. Ukuphinda umba womsebenzisi "ukuthembela" kwi-frontend. Ukuthembela kwizihlanganisi / iinkqubo zokuvelisa idatha, ukuthembela kwidatha, ukuthembela kwizigqibo ezenziweyo.

umthombo: www.habr.com

Yongeza izimvo