abstract

I-IBM Cognos Analytics ifakwe i-tattoo enegama lika-Watson enguqulweni engu-11.2.1. Igama lakhe eligcwele manje I-IBM Cognos Analytics ne-Watson 11.2.1, phambilini eyayaziwa ngokuthi i-IBM Cognos Analytics.  Kepha uphi ngempela lo Watson futhi wenzani?    

Ngamafuphi, u-Watson uletha amakhono okuzisiza afakwe yi-AI. I-“Clippy” yakho entsha, empeleni Umsizi we-AI, inikeza isiqondiso ekulungiseleleni idatha, ekuhlaziyeni nasekudaleni umbiko. I-Watson Moments iyakhala lapho icabanga ukuthi inokuthile okuwusizo yokunikela mayelana nokuhlaziywa kwayo kwedatha. I-Cognos Analytics ne-Watson inikeza ukuzizwisa okuqondisiwe okutolika inhloso yenhlangano futhi kuzisekele ngendlela ephakanyisiwe, okuholela ekuthathweni kwezinqumo okuthuthukisiwe.

Hlangana no-Watson omusha

U-Watson, udokotela wenganekwane owasungulwa uDkt. Arthur Connan Doyle, wadlalela umphenyi u-Sherlock Holmes ifilimu. UWatson, owayefundile futhi ekhaliphile, wayevame ukuqaphela izinto ezisobala futhi abuze imibuzo mayelana nokubonakala kungqubuzana. Amandla akhe okukhipha, nokho, ayengafani nakaHolmes.

Akuyena lo Watson esikhuluma ngaye.  Watson futhi iphrojekthi ye-IBM ye-AI (i-artificial intelligence) eqanjwe ngomsunguli wayo. UWatson wethulwa emhlabeni ngo-2011 njengomncintiswano we-Jeopardy. Ngakho-ke, ezimpandeni zayo, i-Watson iyisistimu yekhompyutha engabuzwa futhi iphendule ngolimi lwemvelo. Kusukela ngaleso sikhathi, ilebula ye-Watson isetshenziswe yi-IBM ezinhlelweni eziningi ezahlukene ezihlobene nokufunda komshini nalokho elikubiza ngokuthi i-AI.  

I-IBM iyagomela, “IBM Watson iyi-AI yebhizinisi. UWatson usiza izinhlangano zibikezele imiphumela yesikhathi esizayo, zenze izinqubo eziyinkimbinkimbi, futhi zandise isikhathi sabasebenzi. " Uma sikhuluma nje, i-Artificial Intelligence iwuhlelo lwekhompyutha olungalingisa ukucabanga noma ukuqonda komuntu. Iningi lalokho okudlula i-AI namuhla empeleni ukuxazulula izinkinga, Ukucutshungulwa Kolimi Lwemvelo (NLP) noma Ukufunda Ngomshini (ML).    

I-IBM inenani lama-software ahlukene izicelo ifakwe nekhono lika-Watson Lokucutshungulwa Kolimi Lwemvelo, ukusesha kanye nokuthatha izinqumo. Lona nguWatson njenge-chatbot esebenzisa i-NLP. Lena indawo eyodwa lapho uWatson enza kahle khona. 

Okwake kwaziwa ngokuthi yi-Cognos BI manje isiphawuliwe I-IBM Cognos Analytics ne-Watson 11.2.1, phambilini eyayaziwa ngokuthi i-IBM Cognos Analytics.    

https://www.ibm.com/common/ssi/ShowDoc.wss?docURL=/common/ssi/rep_ca/4/760/ENUSJP21-0434/index.html&lang=en&request_locale=en

Njengesifinyezo se-ICAW11.2.1FKAICA engaqondakali, 

I-Cognos Analytics ne-Watson yisisombululo sezobuhlakani bebhizinisi esinika amandla abasebenzisi ngamakhono okuzisiza afakwe yi-AI. Isheshisa ukulungiswa kwedatha, ukuhlaziya, nokudalwa kwemibiko. I-Cognos Analytics ne-Watson yenza kube lula ukubona idatha ngeso lengqondo futhi wabelane ngemininingwane engenzeka kuyo yonke inhlangano yakho ukuze kugqugquzelwe izinqumo eziningi eziqhutshwa idatha. Amandla ayo avumela abasebenzisi ukuthi banciphise noma baqede ukungenelela kwe-IT emisebenzini eminingi yangaphambilini, ukuhlinzeka ngezinketho zokuzisiza, ukuthuthukisa ubuchwepheshe bokuhlaziya bebhizinisi, kanye nokwenza izinhlangano zikwazi ukuthatha imininingwane ngempumelelo kakhudlwana.

I-Cognos Analytics ne-Watson inikeza ukuzizwisa okuqondisiwe okutolika inhloso yenhlangano futhi kuzisekele ngendlela ephakanyisiwe, okuholela ekuthathweni kwezinqumo okuthuthukisiwe. Ukwengeza, i-Cognos Analytics ne-Watson ingafakwa endaweni, emafini, noma kokubili.

Uphi u-Watson?

Ayini lawa makhono okuzisiza afakwe nge-AI? Ithini ingxenye yeWatson? Ingxenye kaWatson “iwukuzizwisa okuqondisiwe,” “[ukuhumusha] inhloso yenhlangano,” kanye nokunikeza “indlela ephakanyisiwe.” Lesi isiqalo se-AI — ukuhlanganisa idatha nokwenza izincomo. 

Yini i-Watson futhi yini engekho? Uqala kuphi u-Watson futhi umkhiqizo owawuwaziwa nge-IBM Cognos Analytics uphela? Uma ngikhuluma iqiniso, kunzima ukusho. I-Cognos Analytics "ifakwe" ne-Watson. Akuyona into yokuvula ibhawudi noma into entsha yemenyu. Ayikho inkinobho ye-Watson. I-IBM ithi i-Cognos Analytics, manje njengoba isibhalwe ukuthi i-Watson-powered, izuza kufilosofi yokuklama nokufunda kwenhlangano amanye amayunithi ebhizinisi ngaphakathi kwe-IBM abelokhu ethuthuka.

Uma sekushiwo, i-Watson Studio - umkhiqizo ohlukile onelayisensi - ihlanganisiwe, ukuze, uma isilungisiwe, manje ukwazi ukushumeka izincwadi zokubhalela ezivela ku-Watson Studio zibe yimibiko namadeshibhodi. Lokhu kukuvumela ukuthi usebenzise amandla e-ML, i-SPSS Modeler, ne-AutoAI ukuze uthole ukuhlaziya okuthuthukile nesayensi yedatha.

Ku-Cognos Analytics noWatson, uzothola ithonya likaWatson ku Umsizi we-AI lokho kukuvumela ukuthi ubuze imibuzo futhi uthole imininingwane ngolimi lwemvelo. Umsizi we-AI usebenzisa i-NLM ukuhlaziya imisho, okuhlanganisa uhlelo lolimi, izimpawu zokuloba nesipelingi. Ngithole ukuthi, njenge-Alexa yase-Amazon kanye ne-Apple's Siri, kuyadingeka ukuqamba noma ukuphinda ubeke kabusha umbuzo wakho ukuze ufake umongo ofanele. Ezinye zezenzo umsizi angakusiza ngazo zihlanganisa:

• Phakamisa imibuzo - inikeza uhlu lwemibuzo ngoMbuzo Wolimi Lwemvelo ongayibuza
• Buka imithombo yedatha - ibonisa imithombo yedatha ongakwazi ukuyifinyelela
• Bonisa imininingwane yomthombo wedatha (ikholomu).
• Bonisa abathonya bekholomu - bonisa izinkambu ezithonya umphumela wekholomu yokuqala
• Dala ishadi noma ukubona ngeso - incoma ishadi elifanele noma ukuboniswa kokuhamba phambili kokumela amakholomu amabili, isibonelo
• Dala ideshibhodi - uma kunikezwe umthombo wedatha, yenza lokho kanye
• Izichasiselo zamadeshibhodi nge-Natural Language Generation

Yebo, okunye kwalokhu bekutholakala ku-Cognos Analytics 11.1.0, kodwa ithuthuke kakhulu ku 11.2.0.  

I-Watson iphinde isetshenziswe ngemuva kokuthi “Izinsiza Zokufunda” ekhasini lasekhaya le-Cognos Analytics 11.2.1 elisiza ukusesha izimpahla ku-IBM kanye ne-b.roadumphakathi. 

Ekukhishweni kwe-11.2.0, i-"Watson Moments" iqale ukuvela. I-Watson Moments iwukutholwa okusha kudatha u-Watson “acabanga” ongase ube nentshisekelo kuyo. Ngamanye amazwi, ngenkathi wakha ideshibhodi usebenzisa Umsizi, ingase ithole ukuthi kunenkambu ehlobene naleyo oyibuzile. Ingase inikeze ukubonwa okufanelekile okuqhathanisa lezi zindawo ezimbili. Lokhu kubonakala kuwukuqaliswa kusenesikhathi futhi kuzwakala sengathi kuzoba nentuthuko eyengeziwe kule ndawo maduze nje.

Siphinde sibone u-Watson kumamojula wedatha asizwa yi-AI anezici ezihlakaniphile zokulungiselela idatha. U-Watson usiza ngesinyathelo sokuqala esibalulekile sokuhlanza idatha. Ama-algorithms akusiza ukuthi uthole amathebula ahlobene nokuthi imaphi amathebula angajoyinwa ngokuzenzakalelayo.  

I-IBM ithi ukuthi isizathu esenza sibone u-Watson esihlokweni sesoftware kanye nezici ukuthi "uphawu lwe-IBM Watson lusiza ukucacisa ukuthi into ebalulekile yenziwe kanjani yi-AI."

I-Cognos Analytics no-Watson iboleka emaqenjini ocwaningo kanye ne-IBM Watson Services — imiqondo, uma kungenjalo ikhodi. I-IBM yethula i-Watson cognitive computing ngamavolumu angu-7 nge-Building Cognitive Applications ngochungechunge lwe-IBM Watson Services Redbooks.  Umqulu 1: Ukuqalisa inikeza isingeniso esihle kakhulu ku-Watson kanye nekhompuyutha yokuqonda. Umqulu wokuqala unikeza isingeniso esifundeka kakhulu emlandweni, imiqondo eyisisekelo kanye nezici zekhompuyutha yokuqonda.

Uyini u-Watson?

Ukuze uqonde ukuthi u-Watson uyini, kuyasiza ukubheka izici i-IBM ezihlanganisa ne-AI nezinhlelo zokuqonda kwengqondo.

1. Nweba amakhono omuntu. Abantu bahle ekucabangeni bajule nasekuxazululeni izinkinga eziyinkimbinkimbi; amakhompyutha angcono ekufundeni, ekuhlanganiseni, nasekucubunguleni amanani amakhulu edatha. 
2. Ukusebenzisana kwemvelo.  Ngakho, ukugxila ekuqashelweni nasekucutshungulweni kolimi lwemvelo,
3. Ukufunda ngomshini.  Ngedatha eyengeziwe, izibikezelo, izinqumo noma izincomo zizothuthukiswa.
4. Guqula ngokuhamba kwesikhathi.  Ngokufanayo ne-ML engenhla, ukuzivumelanisa nezimo kumelela izincomo ezithuthukisiwe ezisekelwe kumjikelezo wempendulo wokusebenzisana.

Ekukhulumeni nge-Artificial Intelligence, kunzima ukungakwenzi i-anthropomorphize ubuchwepheshe. Kuyinhloso yokuthuthukisa amasistimu okuqonda anamandla okuqonda, ukucabanga, ukufunda nokusebenzisana. Lesi isiqondiso se-IBM esishiwo. Lindela i-IBM ukuthi ilethe amanye alawa makhono ku-Cognos Analytics njengoba isigqoke uhlobo lwe-Watson.

Akunjalo okuyisisekelo

Siqale lesi sihloko sikhuluma nge-deductive reasoning.  Ukudonswa kwemali ukucabanga umqondo othi “uma-lokhu-ke-lokho” ongenakho ukungaqiniseki. “Nokho, ukucabanga ngendlela efundisayo kuvumela uSherlock [Holmes] ukuthi aveze obala olwazini oluphawuliwe ukuze afinyelele esiphethweni ngezenzakalo ezingazange zibonwe… uyakwazi ukukukhumbula.”

Ngokucabangela ikhono le-IBM Watson ekucabangeni nasekucebeni kwezinto eziyisethenjwa, ngicabanga ukuthi elithi “Sherlock” kungenzeka ukuthi kwakuyigama elifaneleka kakhulu.