g 50, will be treated as malpractice. 2. Any revealing of identification, appeal to evaluator and /or equations written eg, 4248 Important Note 1 On completing your awe. Ddompuloril draw diagonal eros ines onthe remaining edn wsv [oholelalck [ols] : GEOSISCHENE ; © 18CS72 ‘Seventh Semester B.E. Degree Examination, Jan./Feb. 2023 eS ig data Analytics 7 Time: 3brs. »Max. Marks: 100 1 10 Note: Answer any FIVE full questions, choosing ONE full question: from each module. . Module-1 ( a. Define Big data. Explain the classification of Big data, _\ ~~" (10 Marks) b. Define Scalability and its'types along with examples. (>), ” (10 Marks) OR a. Explain the functions of each layer in Big data architecture design with a diagram. (10 Marks) b. Define data preprocessing. Explain in brief the néeds) Sf preprocessing. (10 Marks) a, What is HDFS? Highlight the important design features of the HDFS. (10 Marks) b. Bring out the concepts of the HDFS blo¢k replication with an example. (10 Marks) ‘OR «, a. Explain Apache sqoop import and/éxpott method with neat diagram. (10 Marks) . Demonstrate any six HBase commas with output: (10 Marks) Module-3 ©). a, Explain about NOSQL data Sf ‘and its characteristic’. (10 Marks) b. Describe the principle of working of the CAP theorem. (10 Marks) OR ee a, Demonstrate the working of key-value store with an example.(~~ (10 Marks) b. Describe the features of MangoDB, and its industrial application. > (10 Marks) . Module-4 a. Describe the Map tasks, Reduce tasks and Map Reduce Execution process) (10 Marks) b. Describe tiie Hive architecture and its characteristics, >) ao Marks) ii OR y a. Demonstie the pig architecture for scripts data flow and processing (10 Marks) b. Differentiate between Pig and Map Reduce, give industrial applications for each. (0 Mars) & Module-5 Pes a-—-Explain the simple linear regression analysis. (Loria) b. Demonstrate frequent item set mining and association rule mining. (10 Marks) OR a, With a neat diagram, write the steps in K-means clustering, (10 Marks) b. Explain the purpose of web usage analytics and the significance of web graphs. (10 Marks) ween1a. Karnactat, Law Co ciety's vichwanatsrao vethpande Ine hhete oF Tech nwloryy, Haltyal - $8 1324 Xepacl ment 2 compte, Sdtnce & Engineering vTU @P Loluk¥on. Enbject + Beg Cale Arulyhat [ woot] Yeas of Exam : Jan [feb 2023 Name ay the Pout ry > Profs Yaemeccs Chater MocLule ~4 ote tine Big deda- Explain He clacct hrcctliod 2] Big dala. [ropa Bra dda is Abgh- volume , Kigh - velect ty high - variety in penton ane thal and fox requints ew foume 9} prove sting ia enhaned dectcion making , te cfgkt Ate coven curl pre cere opkimigatluin - the claxtlecaltin 3 Ba dala ase ae jolers f) haty fovicet ~ bala Shorage uch as records , Roem, atch butid data bares , yor Oehented Tn- memory dela table, column . oMentid Tr - me clota stables , data worerowe eet W) soota foemmtt - Shrechued ¢ demtchuchwed iw” Bey ala foutie = ata shorage pec, ove, cledamarkr dette wocrehoure 7, MeLak Alb , fendoes , exteenal dalla ,weathu ceca et. Prof. Yasmeen. Shafth, Deptetcse, KLIS vor, taltyal. Jfy) og data fonecle - uvduchued, seat chuichued ¢ mut} -chuchued dah, vjbata Shr Chuechive ~ heb jerdespnce pt Cloud setvery data barehourr | geoph Alb ele Ww) processing dala vale - Batch , neat- Lin’ » eal - Taw chreants vii) Provexing Big Asta Latur - Highlume , Vveloctly vanety f retacéhy , batch, neat Leal lie Pro cexcing Vi) Analyse hyper - Batch , cdedubed , neat seal ne , dokn cer analy Ke im) BFG kata, proweaing — Batch protesting , veal “Lente methods Pre coring w)oata analyse — stalichecal analatio , predichre method analysts , Legnescion analycty, me alge, feat analysis ee ye) Bata, meager ~ Human, dusinex proce , browhedy « cUicconenyy pesapriie AppWcalienr , state thes. Ie: elie ccatabttity and Tu fYpes along wh ts er anepler « Cora) 2? Seodal tuby enabler facreaue or decyeare the capat ty vf date Sbrage , proce xin) and analy BOP. Ccotabtlety Te the capabPltty sf He yen to handle dhe vwostload ai pe the magnitucle af he Work eygtem, capability increaied § Kovkleadss needs increment Rith the Prot. aineon Shatich deph of coe, RUC verr, tHatliyalWhen Fhe woklad and Couplxity — exceed He cyclen Capacity » cele Fl up g sate 34 ond *) Ves h'cat 5 calabPlity Vertrcat wcalabilly — means scaling w he sytem — nagourcer Aud 10 Creating sthe Syelem Laat. , repertteg — Argh Visualinatiss Capabflilie . - TAS U an addi bboal “ary bh wolw problint 4 quant lonapleniteer - ~ Staking wp Means deri gnuig the algor fi according do the Gachttechire At tue woour ees afpicie aly» mr a deaby ty deta tte . fn 4 poe Procercg , Code C19 bith Ta cre ang Comply tocrease by factor 0, then Sealey Up nent hak proceroiry dake equal, lese fran C nxt). Wiserionetat ecalmisihy racreasing the Hori goat cabal tile, meant : nunbee af cyttenn wovkeng i) Conherente And ccolig ont dhe workload . - Proce sting dfn dataceh 4 large date cel deploys Rocizontal ecalab?lity . ~ Scaling ont Meant wlihy mote peroutced — aval Areytorulog doe proce xy and chrage toby 19 allel , m x tra byte hy reoures AO Sytem proce of daw wh tie t, Men CPx a) tra bytes perce on Pp paralret Achibnted rccdus ouch Malvinas 3 Prof. Yaseen. ohana, Devt of (SE kit vorrHat Unie taken up remaind lightly more Fe oh. Explato We quoition: 0) each layer wm Bi dat, drchitechiw design wiih a cltageau Bom « > [Laser sus] expos of Bade tages] Panay Re Creal Fi Data cel arc , BLS, Ines seal-limad Frodedge cetoveng| | pothing vious tifa fencer cloud , web ete Cynchronous acy nchrongue ext fjocar Aocka cal A [eer beat Laeaph Ab , Case ardy ribnae db Prasng me, | Tdents Keahin af tntanal &Exksnal Sources Of dade i) Leyes J is 4 Fobernal q& booth * Li conaiclee — felloroieg apes th a denigo, - anount date needed at 2, pwoh frm Li o¢ pull byl? , _ source lata Ayper db; ftles , Web oe service ~ Soucce formal - Ceui = Ghrechured , uattnec huned , Sruchired, V)ba Fe for data ngeshod Le donsidere . L b ngeshdo ¢ ETL Processes efticr 5 eal tue ? phtch nuean Shore and fue the clat, at Jenralid oe i batchy. Shatoh, dept of Cee, KU Voir, Haliyat entecnal , ot tdenbhyeg data Counce, Shich Ate Pref Nosnmeenji)ta pe jor Shrage af data dow the 12 layer , DH contre - bata Shrage Nge ¢ feemat , Conepreicion incoming data brequaay , qua yey patterns ¢ ortunplon equive ment joe le e tC ~ bile fioege why Des e¢ woceL dala shored. ie te date Pro cenit tag Kethoue @uch as MR, He | Prq 8 party TH conitdeyc = eOatn. Pre cerning Cofhoarts Such as Me, fine , pig Spade » © - Pov ceating w) scheduled — bescher /raltiai oo hybid, ~ Proce Ar pet Kyrchronovs et oynchronovs wequitient, at Lo VLE Fe fee dal tonrursplis’s HH considers the Contumplioy ef dite for the Jotosag bata rnrregeali9 ~ Batol wages doc reprhing 4 vicualizccle, ~ Analy bu 7 BPe, Bes “Knowledge clit Cove ny ~ Expook Bp dela set fo cloud, meb ar other wyctemu . 2b. sbeftoe data preprocessing» Explain ip brief the neck 8 Pre proceaé ings [10m] oaln Pre - prscescing ic GD ienportand Gtep act the ingethon layer . Pre pro te ccing fe a must be yore clata mining and Gnarly sa Pre proce scing Me algoridéon - Analy Kel heeck poe ecreenirg of dake quality also . te ake a murl be yore ruenniig a Prot Yosmeen. hatch, . pepiotcec, els voIT Hallyaledata when berg “ported fo a cloud vers ce oe clotu Shore need Pre~-proceacig . Pre -proce s£iny need are *) Prop pag ont of oange , inconuteterd owl & vale: Ww Per Heung unecltable, iovelevad ¢ redundant informatie y iv) obata ecleanenq , edilig prducte's aad Or oraaglerg: w) bata valtdalier ) parafer altos K toanstodiny v) ett provercing > i} ng - velete fe role ef ve MOV! oc TS apl ye tratcurale a relat party of the daka ayict detecting thom . ®) date cleansg hol - oate cltaniy u done bejpre Mantixy rt data 5 Sncwmpletr ot avrelev ant date pay srondt Foto rate leadbiog der{etoms' Ti Cah Envichmect —refue fo opetaliim @& Prtenses Wich Tofu, emkwer OC Tapront the ras dala ©) sheila gating -yofere A prten reviewing & adjuslag ype acquired datacel ») sbato Reduction — enables the transfeemeation ot acquicd DRemalts, Yo an eedered, Cocrect ond siuplthed feu WD obata woracajliny —yefeer te the procen af transfoe ting ant varied the clate - Resnick Oa) avalyur Gre then Apgreprial anol valuable. Pan armen. Shobha Dept of SE, kus vorr HaliyatModule 2 Sa Whe Ue FDR. 2 Hegiligh the ion portant ales'ign Featiues 9 HDFS [lor > HPS w a sytem of Shoring files at declvibuted compulig nods aceoccleng dy tHadloop auckttechire and accesstotlity of data “blockd aytec jrodug aitevene fo He rack « chuler ~ HpPe pas designed yor BD Preceating J up pou many wen Cfo ultaneovely « Tupertant Pengo feahues 9 HDPE - Hope te nod a hve peated pile gyetem , the deign agcumes a laye file bovide one [read aoa nsdel track enabler BAno ephimnizatio ar + melon many of he tonumenty & Cohfitence equi rementy - The durqn 5 Hor & based on the design GEL Gpoga fu Sytem? « ~ Hors ce aleargned fer Aata Sfhreamudg POKER E lange amourti of dota ate tead from dete ra blk — Tuport ant chenrgir apect 4 HOPe tc data locality - Whe torprtaltoy ue aowd & dala sot, movi date to the @nuputahoo — H9FS dos redundant Aesiqo Ret Can folorale Sytem failure ¢ CHU provide the dala needed by the couprt: past 4 he prego Prt. VYaoseen. Shathh. Pept of SE okie vonr HalruatThe Aeiqn % wpps te bared on tao hyper ob udus 7) Namen 1) Mabhyle ott rocles 7) Stagle Mame rode manager att the metahata needed te Lyre and retiere fhe achal data fren the — dldanodas« PMedata pnd are ~weponsthle foe ferwiiy aad grette Tank thm the fil Kyctim fe he client - The Hope dango & a aaaoke [clan axchitechire iy WhEch Reader CName weds) Manager the pile Sylow Amen pace & segilals access yi ba cient . - Name aode Manages block cabs, , debts, + raplecdles, - ~ bata rede peje atral proceraing of dala : Bb Bring ont He Concepta He Hope block replécabts KAA ay Chample Cio”). > When Apps Write aye , Hie reple cated alm sine cuuko « Foc Madoop olurkrr fontainy mace thar § dadanedas the sepieentiion vole 1s oe. — Hors defautt bleh ay ws 64m8. Tog robe ye UG wiHen fp HPI, vt wi deal, Uock tat “U apprevimately 2ORB Wy SIX* ~~ Ha jie oH Se Come te rotiHen de HDPE, Pst. Yasmeen. Shates . Dept of CSE, RLS VorT, Haliyata 644 blot 4 1646 blech pill be croaked . Pique below showe hor a file U bye ben Molo blocks 4 rep ecabed acmes the cleats. me 3\4 S| patapl - 1 (3) @ a ode patanecls 2) | a a) iS Jo HAS Core, 4 veptecalion fades of 3 me “that any One drdanode can fatl & tae replies blebs well be av allele 9 SAU noda, 4 then Salcequety yo-repltcobed on Otte lata noduss 4ay Explain Apache Sqpep (Fo port and tupoat method WEA peat Abagean . Ctom] if > Apacs sqoopr > Pr The data enpoot tc done > Sep DIN the Piet Shep, sqeop erameny the database to gathec the MeCL ney wetadata for the dak be be ia gored - 2) The getod sep fe a rraup -onlig Hadoop job Taat Sgoop submit te Be clutter | TAU “job doe tHe date, trantks tiny Medadota Capatered — 1 Prob. Varasoo. Chatha ~ nevbebror. kit WIT Havel 4 requprviour ore - the tiuported date Gr waved wy an HDEE Aine Fegure betow tllubrales the pro cea Eg yore et ge qaese | $4687 job HORE Sher age > ' 1 (00 Gather metadale | Map ! { > 1 1 } 1 Apache £qeop Expert (2) Srbnat MAP—Onley To isis ie T HDPE chi die ‘ ‘ Ww Gattor methdada | | Le \ ’ ponte} Prof. Yatmeen. Shai, opal Coe | AL vor, HaliualFigur abore Shows Apache —£G00P orport nedted . The erport Ye done Wo) Steps. ) diet step ueodo Cranty tre Aatabage jor hurtadocke a> tapes! Bep Wel Map ~erly Hadeop Job fe pit the data do the catabar . te Joput dat, get ht? Spf , then -uler Tadlividual aap talks bo puh dhe Splits to the database. Ab. demon brake any x Hae comrmanclr iSiHh or [10m ->f) treats the database hbase Cmatr) 1006 :0> Creal ‘apple |, ‘puce’, tyolume | OAnsé(G) 1A O-8ISD Seconds - +> fable Name u apple 4 do columns are dep ied i) Pot Conemancd, ey wed fo add dota t He price & volume databaye fren pithin Ke shel. Put ‘apple! ,"6 -May-22'," price topen '," 1225" por ‘app’, "6 -May-22', ‘ Pree: high , 26-46" rae ‘apple ', | 6-May -22 1" Pwce: low', '123-36? wt Vas a i . , P ones 6 Muy 22, price 2 Close |, rol pt ‘apple , ‘6-may-22',' Volume nt ur0zeP! Pref. Yasmeen. Shaith, pepret sey Kut von, ttaliyel wvHii) Dnepect tee __datnbase ; te ete database Can be treted. tat ng Ye Sn Command , Sean ‘apple’ hbase (mais) 2006: 0> Scan ‘apple’ Row cet 6-raay-22 Column =price Close, value = htg-o | 6 tay -2 2 column aprice thtgb, valyes [26° tS 6 May -22 Columa «yrce: low value > 123-34 Eo Paay-22 Column =prce open , Valve = 125+ 25 6 -May-22 Colume zvolume ?, value = 410 2038F Iv) Get a Row Nou (an we the mn by fo Beco an individual Sows. hbaseCman )+ O08 : O> ged ‘apple’, 6 - may-22 Colusy ce Pw Ce :clote roertamp = 14305462 7 Nalwe = adel Price : hig sh, Pranep = LBC S4AbG 1 Nalve , (26-40 Pace: low eresstamp 2 lazowalt , Value - 123.394 Price + Of bende Shamp = Hee SAE Nalwe ~ tar ag volume Hin ictamy = ItaesmP, Value - #820324 v) Ddeleke a Pow hbase (maty 22004 s0> delebr aur “apple's ©6 May 2-2" WO Romeve atabl TH Minerva o& tabk , bE murt fixer disale Ft, hhate (matin) :004:0 > disable ‘apple’ hbate Lmain): O10: 0% Arp ‘apple’. Pref Yaseen. Shaith, Cephefcse, Rue very HalyalModule - 3° Sa E xp (aun about woe dale Shore ancl rd charactortehy - [ror] > NOL dat Glores Me Contdered ay demi —chuchwed date . WOOL DB doe nol waqvire specialiyedl Roemc ke Shrage 4+ hardware fe proteraing , Shorage can be on © Cloudy big bata ore wel HOLOL » nosoc dats Shre Charactercthu are HP NOLOL YW a clan % hoo~ relations dole ehrage wyetem tty fe atbe data model. ay Wocel rot necerartly boo a fired wala thre Gg thu faut pleeaal 1) Hpaches HBase - Hors Compaty bu , open-coucte , 000 Aeliliond datachre -potfen inJova . providua scalabiluty p ewal Fency « W) Apache MongobB = HDFS compalbh mares clare modtt, clocumert otitnled opn~souree NOL scalebk 4 ner ~relahonal Alb . Wi) Apache Catcanle - fore compakbe pg decenhaliged Pees fy pees , Nol al, opencource fy “CA. are Py Atleacl hwo are prroert he applscalisy [cervite | mover - Conatiskncy Means all copter have the Cane Valve Like tyadéhone! PBe . Avaflabrlity means atleacs one copy case a partsKax bewmes wHachye ec atu. packttson means pact Iohich Ate achne bat may net cooperate ao in Aichibuted dBc. 1) Coastelmcy ws the aichibulid — databacvs mtann all the nodes obser the fanu dal, at he m avatlabe Cone tine - ae operatons WW One pasbrhon of the databace Groth rotted 1h otter gelaled parhhong cm cme of dirchyPontiid ab. iD Availabilty wean wat durving fhe transac doo, Fhe fret values must be avatlabk 1 0 thes pacsisonr tt the dedabre Go theft each requeot weretres A retpende va Succen Ao helt as fatlure . i) pathhoo auane Wvicwn tha lage Aatabase (ats dif databases tothont affecting Whe Pat Yasreen. Shateh, Deptetcse, Ki vorr Haltyatopsation 09 “hen by acepling cpecthed precectuures fsffhtn _folecanes - refers fo teoKnralie, of 9 Lalor, gia Woe ern 9 Core vl Menage ese, nocle Failure ec rede rot weachabl » 6A Demonstrate the noorteny of key value fore pith an eranple- (rom) —> The efrpleat ray do fap [enced a Schna- les ata Shre fo te We frey.vale par. - ata webieval to fast fo tey- vale pates Aate shore. ~ A cape fhing cated , key aps Fa large data «hei er BLOBC Basie Lage objet). = key Nolve Bhor acess we a petouasy, kecy foe ocee King He Valves - ~ pfq howe fey Value pais avchitechual patteeo Bo eet nash | cade clay *8' i [Malou | Branh + 6299 ” cee 2 chudeet j ‘alue co a oy bey vi : ie 0) deta Sieve tan Che any oda “ype mH aN® jeld n) A Query ogquert He Value broherne He Valve Qa single ideon i) Pey-Value shore consistend - Prof. Yanweer, Shailesh) Deph of Cée, ke woaT, Haliyal je‘ iv) by - valve clatn pve may be hferanche cal er may be tedured bey wale clon - v) Rehm Valuer on queiz Can be wed fo Gnnect foto lik, table -Columm , data -frane jids g Olumns- Vi) Hove ftababflity , reltabtltly , perdalbt Ey ¢ low operakional coet . vit) ey cande Lynthobee’ OF duths - genucalid . The buj-value shore provider clind fp read + weft, — valugs Ui @ by As Jellowe : 1) Get Cbey) — re heen he vale amoctalid NitK ou fay it) prt Chey ralve) ~ aroctate tre Valve i He bey & updedles a value Cf this key already — present iii) muget Clegt , reys, “= key 0), poderns Fae (itt of , Nalves aasoctated tot the Octo} W) salt bay) © prne a bey tit vale Rom the data Clow. rhiuttations 7) Wo dents ane Main fared on values, Cobcet of Nalue G not Cearchable » i) key-value Shore dott not provide tradtbonai Alb Ke pater bind a Ahutaty , owircbeny AN NAintaenstg unbave valuy ae hey ¢ rote difuutt solen he Volune Foahnu 3 PAongeo DB Ape i) Mongppe data Shore ka phayct cal ton teuret collection, . Each DB ge tk Onn cet ¥ {ile og tA yu cyte - A numbec 4 DBs Con BA OLA Cingle raged deoven ii) DeshSbided PB maker avai abotlity Ligh and Provider — hotizentat Scalability ww) Ahautc epeeations on a eiogfe Anuumet Can be perjoned even Shough Support 9 murht docunerd trawacdting te not preven - W) obeop quay absbtty - cuppork Ayame quer ® Aourmond watng a cotumedl ~ baud quay language that Ua porrttag ae £OC- v) alo Conjtquialae Cache - RorgodB les AM free Rrenrouy 69 the Suplean auborvakicably by hay wemouy ~ Mapped ye Th mo wed dake ‘uo bepth Ram. SH wtexes axe outbid pe quener & tee working datact ft Ram, Mmgno® fetveg alt queries fon meMary , V) Queuying a 4 heal tina ay qcegations alton acering 4 Bat 4 3 he | Aut, pptetendlay : vil) ofbcinent rrodel a welt Atpiord. dbtunerdd wee TOON ox ebnrding dota. ouverts here Aynance Schema - Prt Naimees- Shatety ) Dept vf CSE, RLS DTT, Haliyar.Wii) cottecttcn —ghores A Nembee a M1onqo DB dowmint - a4 may ohre document = FAact do pat hare Carte felts. Thus dp comercy 9 Collects oe source Ler yo) Monge d® “gy a clocument — hore tH Which — one fotection Rots cliffetant — Alecunmeonti - ata Ohve Uo he fom B Toon chile clolu ments « umber 4 pielde ntentr ¢ erg 3 Tac Aoumeont Cay Arfter fom ene Aowrmed th ancthte + x) Skeing vf dale ve }atbh + daly dbre congixla 9 Toon (eke doument Thu implies Yat fe fields aan vary — gen dotunest do document & dade thet can be Changed over Lend Pree Yamecn. Shobth Dept of CSE, kLe VOI, HealigalModule 4 fa. Weecrtbe Re taret , Kecduce taste. and Map Reduce Exel process - [liom] 2? ap Tatts = Map tuk Means a task tat Yonplunresti a Maped , which mint wer application Cote ty each oy -Nale pare CKyvi. Rey BH maps 9 Geoup 9 data Valu Vi. ~The oubput vy Map©) would be zero oe fotumediate bey value pairs Ck2 v2). Keduwe dash fede fork refer to. a dash yohteh tater dhe butpel v2 from dhe map a an Fopnt $ Combines those data pieces Hd A Smaller ¢ ef oj data Weg conrbings . ‘Map Reduce exeuttiog prover Trpet Blu [Pap reat; eee | [Map F raat rater PP) Sgvent Hf ovstpat 5 u of rhaprast, bata, sap | ee | pI 1 i MAP gee [Hi Toktrockurs | Leegmust > | anh My i Too | : Maprok | ie | eee Prof. asmecn. Shatte wept of cs, Kile vott, Halyal FIThe figure Obave fhows the. ap Reduce aoualihy precens The Hadeop Jeb clrent gubsnuite Fhe ob amet toodiquaaliéy te the Jab Traber, rohich “Hen AKU the weg pone lity of dichibubing We fet hvate / config uralesy ch the Slaves by g tacks , noni tor ing them Proves drab 4 Dagrosh ( pornealin’ fo the Tob - clienh. - Map Rodirce mink 9 Liagle Manker Jobhacbu g One clave pen clatter ackle Marler U reg pons Bu rad ochodbedlerg Fools 15 a bb onty claves , noni Iovng & wy exeuding the failed acbt . > The clever exeutle the tacks ad. clacectih by maser - Eb pencribe dhe Hive ayehtvectwe 4 Fle Chasack sicheu tom] aa Prague below Chows the hie anchtechie c [Reb Bosces] [usecrpyicae] [ptcalos comprrent: 9 Htre agchilechure are Dt Leeve CTAGAH) — optional Cervice tral allow’ A not cle do cubuit PrtNasmem. Shaith , Deph of Ue, Le vorry Haiyalaequeaks & retrere — reenbde. W He CLP — popular frjerfate > thderack widh Huw. Wi) Web antectace — dere can be ascened tutu Web bronvee. This requing Huot Scorer ; rrennuy ba Lente designated rode, 1V) Meta shore ~ TH i Fhe cyctem Catalog 7 AL other (omporsal tft’ istecact AW the Metachve . V) Hue etree - a4 Manages the Lite Uy 4 Hwe OL Statement Aung Coupe latic', 7 opbmnizahs's 4 exeurlion i) eapabicty — fo tranglate queuit tobe magteduce yobs: Thin rade Hind Figure below shows Pig aachtjechine jee gon pk dala flow an prow 1 the The three Kap be taende LCApK ane Grunt Shell 2 40 nleathre Shell 3 Ng that exenitur the Scop 25 £eatpt Files Py Commands — tition rm a Sevph pul thed exeete oo Pg Sen ree, 5 Prk Naanscen that int ne > ‘ « Haltenl spre pavivonmedD Exrtbedded Lofpt: coente YPPE fot the poncler'no avotlabk ‘ws Pig bt + -c9 opera, LOPE Can be in cir prgtanmdy languages Script] Mepbedute Process | az Mappeluce P es Reaver Enyne + & 7 sain Kanye | pasteo- hk pauer handler Pig oe TE aes — nh Pp a CLR . yndats OlP ua Dae: bytiniiaes ~The pag lu ebuild fo the tages optimizer . The @ptiniali, actiitces Such ou ¢plt, meget , poncho And reorder operators omeude iD dade phoue » @ugi lee — tre Comptlic omptlee atte he epltnigativy Procent. “The optimvjed coco ee a Cerier ot Mapkediece "Jobe . Exeuitivon Engen a Binatly tne Mag Reclure oer Submit Poe exeutlisy of He tagtie . “The Maptidute ‘joo exench and tt oxtpli he féoar reat) . Prat, Yarmecn. chaibs, Depot ere, Ris WORT Haliyot#t) cotHerentia between Pf | inducbfal applic Pig NA date fow language 2) igh tel taguage 4 | ftey Pole Peekocmun yes ' fetter, Sexhey | i ddaue cpeation are | 4 Posgpanunes with a baw Knowledge 4 COC Cay boo Sinuple. Convers entty , | We maulls - queue Opprac, 4 treby redutting He long tt PE the Cody signiReantly « S Ho need doe conptlaticr fe tx euukion , operctlee convert inderna ly indo Maphiduce jot YH Provider Nested habs hye, Ath hep, Boge ¢ maps | pocelirn, 4 manp Reduce + Phdwubtal alton ig cath. ——_t_,— G4 anol Papkeduce , give [Maple Ig Acta preening paradiqn I> low lenebs language are vigtd ’ be taeda aifftcult to pertoon yo, pte, Seah) order 4. opecationr behveay dataset . A} Coup. Java, Tuplemanta ~ Hons requre erpoeare f “Java Language . & Requv® wot 20 Daa mete Fac humbe 9 lee fo Petform Tae Cane teak % ong Compt lerlioe prcex pc “Jobs. TB No suck Aacta hyper ~ TY iotoperaky makes Herm propotel meckaalens fo E-Commerce buy ticshry , wee jntecactun | - Map Redirce | doda. warehouses | Prof Mame Chafth, Dventonis , Channing Cane twued 16 laege Pert of coe. webcile rwords , by ‘ volume dela iy 2a kur wets Hehiuat_Applecation > Pa i proven if (aege otume of data fom he decasels Dwell at every plate whe — analytical inchghty axe Lequieel pity he we sf Cauplerg 3> Poe the coUeckror, 4 lage anout of dadued in the Fem Q web Cramlr efnear. Leqremion fr & Spe onal tordely vase algovisamn é ot ud Supevitedt ML algevitom fy. predchvé Enalys of “pt edly 0 Yelation chip — behverr? be Inelepentls predicla 0 caplanatiy — ane she deperclent duh Come me vous ,y tmiy Oo Uneantty equalio, - Ye $ (aoa = 44,7: ihre a, ad a lonstont a, Ww the Lineasily Coe i Cleat Sipe Ginens keqywarion FF purfoued when the requirement poedtcles, of value of onc vartaby, pith girin value 9, anctir varalir: tH contd a Cample of (0 Shideats Pox thon pice GPa’ ound Ragh School Sores (HEP. Ae fCnn1w9 - Pre. Yasmeens thatth, vephof coe, kis vorr HabyalAprume Linear Legredion . Theo Gpa = by. nep+y figure fhowe & Limp Lirta Leg rer ple F foc the Ae cbin tap behoeen the Collige GPA # He percentage 6} high Sdwol rmarb. — Plot he valuea of a geaph , with hugh Cehoos Cortes Ww percentage 09 He 1 ari and PA 69 He ams » foro Geo 20 ts eo Phin res tae peapeet” Linear .relalidy Shep bakveer QPr & Akigh cchool eure nisi, ate 10 poids 09 the qtapr wonld jit coa thacght dite. ~ vohoneres a emperfere — Benoa Aelalis's chap tists bebween true too vartabls , a cludes “of point en we goaph WOAH Sepe. Mp waed 1 aay be Obtathed. ~The pespore 4h Regreicwn Analya uo come “p Neth an equalto q 6 due that jek Brough a cluster ot pork HK aad acad amout 9 devel’, jee tae At. Prt. Yasmeen. Shalit, Dept efcee, Kis vorr,HaliyalThe — bet — filled leat Ui called fhe Way dui. » (+-——_—— g 8 a e ‘ a Kp » HEP OSC +4 a = B- GPA = HEPKO OE 4 SOL oa t higwe Sonn’ a Civple Lindae segrekan Pi fwo regret 00 Luts wrth aiffe rear pegrea bon equations . - The toodiner oan ft bert fe Summanye the elalid, bebvern GPA ancl hégh Sdwol percenlieo ge * Ber ralinny mieeeced, | yucca oe od sf eater eshrualion Anche & 4a =a, +4) Qi Qe Hi" hem. do da, Mt Good ficunte . er a Cllet - Prof. Yacmeen. SAaibh, bept-ef Cee) Fu vot, Hel watab} Berons bate prequent Hem. Set anton ng and ae clalioy mule Masing . ftom ? Frequert — Penaced fatning CREM): - Fim vofese to a dale Mining method ohich hetpr in developing the Wenuetr taf appear breqverdtey in a clatacel » seramp - puddin a cel of Shedinly ho “pnag vendly Choo post performance wy Semysfer eranrinakons, — Frequent Lubgequvene. ti A Ceqvente of pecttens tal occurs frequent y : aA - porehoxing Q pootbat follows purchasmb of Sporls bit. - Frequent Suhchechire refers to dif'Runt — frechutal Roms seth at gfaphs , fees oe lathes | wObue way be conbenid with Btensety 64 Subasequen ces. ~ Prm uw one Of she fopnlar ecAnequer fp valrade Faonledge From data. - ~ Pra pends fre reg wba’ lees ec) we data - ~ Fam fs We preceduy dep fe the arnoutalts, de teatning agov tina : = Oh Sater aalyiy reowtk to 1) Tecpro vencent oh Arrangenunt ©) productr in sholves a) Bling 4 Sale — peomohor ili) Planneda A, preducd tant a shre Showd Srck wy? VY Support Brot - Celle $ Prduct budbig. Prof. Yasmern. Chaite, Dept” ob Cee. kee oer PAbuet 24Wt) Ascoutales, Rule mining - -f+a lapovtant nmtked. ef data mining - The nutheds fas been htelolig ued Manip Applecals, Oxens qe dictovetig Potenea bag wUation ches —tohich ate prearnt” 9 lage datncets . ~The Objective Ut Yd unowred telaltsnsheps using some Shesog mules « the -tulu are fetrud as anotalto, TH jor fugue “Henaets - the attoutate, ne problem iatutu tart det = (432 .-- 1d? be a cet of dichacl atiientu ¢ Qho Cand Urteat oe Taf dita, --- tnt be act ot n transachor, fF tontaty a seb ef tem Such Wat wer red - An awotrali, rk Y 4a iugle cation of Fe fom, ® Whe XY belong & cet of Okay catud rteawets . OY CT) anh ve and Y ate cligomt Ftmurk Cx nv = 6). - Fence xX LS Catted Cinteteclent, YU lonsequedt - — Arlette, anayss 9S apple cable fo evetah dowatin, wth 0s Masteleny , brovd formstece , rel Ming | cclenti{e clata analyst amd th hruatoy Atichdy Sytens « Prot. Nammeen. Shaeh, Dept ef cre , Kle votr Haliyntloay BIB a Ana dagen, mils he SHpE k-Meang Cbuser ina - { tom) > Top Aate Points Citron Clus tering dinchions Jerakive shes 20K) for [AT the fe Crifenon & clusters Sakicher | ech Pes e Figure Ghows he Sheps to K-meaag clusdeving K meant can be exerted io the Jolloroing pe te Randowly tnibaley the K elustler Cenbord poor (e1,¢2,.-..cp) tu pakbro Centers KER Mean Pak one with eve Cluskr Centric 2 Go Hugh each ot te dale Pointe tnd Aecign Pons a clus whee the dictante fom a Cenheel & wdnimun - 3+ Tends, the contd ot the pew ebuter formed , PEW averoge of ate the data poisk to aclutfer, Pead. Pref Madmen. Shaihh , beph of coe, Ru verr Halgeh4. The pew ropealid unl ne cduage % the Cluster tabu plate S. Gleps 9, Healers — elocabeg —algorrtws 1) Inpet 2 W objets & K ao of clus tert n) Dretpwt > bet of RK Cluskos , bch Wer cottesian Junclises #tk)- 18) Atgonhan Steps * &) wibaly Bh terberde ae the iba golule, b) (Re) conprte — mennber “hip < for te objec tug the cluster Centrords c) Updati enterid vt the clude: ACtordiag to rN Monabeuheps of He checks : A) epeat bem chy CH) unk! Bae or ne Obyeck Change » ia the Clutter Cenproid Lbby Expla tae purpote rh web Wane analy bts and she signi heante- oy} heb qeaphs : [tom heb lage wing — Abc Covers amd analyces he potteens fo cll Ghreany + _ TF also \ncluda awoctatid data geneoltd + Collected at & comequenct ef wer inferactian pth Web YAouttel s ~The 4dhre phares Web cnage me re wrepesye mn 2. Pactttin abisvonery 3- Paden baalyser Prt. Yaameen. Shateh, dept-of Cee, kU VERT Hadyarte Pre procenaing - Converti the Usage ropoemectiar, colletid fem Be varion data Spee tht de clatr Abs hactiony ecettang foe pats ctiecoreey, t Poctteen ~ distoreng. ~ Zoploitt Methods and. algorithm Aaveloped form prelolr , ouch a chebecho , dala mini , ML anol pattern vetog nihéy . WW) Pateen aaalyps’y ~ file ont Un Pate rvaliny subse pacttennn jon she get jound dung the pelteen abit voveny. phaoe . Sigitsane oF Neb _quarpl = Be Keb Glaph model the heb paqe — Page hy pecluiler Ae the property et the gaph wwde - Aecume A page Pq cv? % —linka fron pamur ¢ pain) bur —-(inting cial be Pq) 5 ts taken Nop (Pg WT - pager: ~ page Kank Canbe geen as a moculey a randoo Suder saat Gasts 09 A vanloor page Are Hen at each pot ¢ _ Ev) models ‘the probabel (ity that 2 random Ling Junpe 9nd loanect beh on-Link fo Pov ee) — Rev) meculs the probabh lity Rit the vandlom Rind Connects te pgy at any gen hi — the addfhey sp Cv) dolves the problem of Pav by thane ont Cintug doa (66 pi dead Prof. Yasmeey. Shatth, Cept of cee, kU vorr -Haliyal 92end. - Efeds Oh a lok fpam can be outlrhed tuneg the teplé Centbve Page Rank akyorPan - ~ Unk £paen tig fo mitkad dhe page fank lo tuelde) A Une 4paro —actfemupk fo nuke Page bane agers wn Geapedine - ~ The Span aac tg peqer comet the page crepeatealy ane iocreater the in~ degree of a Page, Shereby enhan cusy the dink fo & lage valve Faculhy Naueg Stqnahuve PreYoumecn . Sharer Gees DeanAcadentes nics