Luận án Tiếp cận ngẫu nhiên trong khám phá tri thức

Minh ho~:Xay dt;lngcaemodunkhamphalu~tke'thQ'pchoM thO'ngd~thangtren m~ng Internet VII. MINH HOi}:xl Y DTjNGCAC MO DUNKHAM PHA "K "K ~, " LUi}T KET H<)PCHO HE THONG Di}T HANG TREN M;}NGINTERNET Ph~nayhQcvienseapdl;lngmohlnhkie"ntrucph~nm~ma trend~thvc hi~nvi~cHmcaclu~tke"th<;1ptrenInternet.Vi~ctimcaelu~tke'thQ'pthvc rachila mi)tchucDangtrongtoaDhi)h~tho'ngd~thangmahQcvienda thvchi~nhoanchlnh. Tuy nhien,d~t~ptrungchovi~cminhho? mining data,hQcvien chi gild thichcacph~nlien quailde"nph~ndataminingma thai.D~tlmhi~utrQnv~nungdt;mg,xinxem[16]. VII.I. SOLU'(1CVE Hit THONG DON D~T HANG vA CHUC NANG DATA MINING TRONG Hit THONG CongtyFlyingFoodGroupla mQtcongtychuyenclingca.'pcacsanphfimv~ an u6ngchocac hanghangkhangaMy. Khach hangd?t hangcua congty tli' nhi~undi tren ml'ocMy. Ph~nm~md?t hangcua cong ty co ten Streamlinedadu<;1cphattri~nbaicongty PSV la ndicongtaccuahQcvien. H~th6ngnaysli'dl;lngcdsa dli li~utrenn~nSQL server6.5du<;1cthie"tke" nhu mQtling dl;lngKhach/chuthongthuong. Khach hang ke"tn6i VaGh~ th6ngbdi m?ngLAN hayb~ngduongdi~ntho?i. Trang56 Minhho~:Xaydt!ngcaemodunkhamphalu~itke'thQ'pchoM th6ngd~thimgtrenm~ng Internet HQcVieDdil dl/aVaGmo hlnh4 lOpd~phattri~nmQth<$thO'ngmoi thl/chi<$n dftydu cac chilc Dangcua h<$thO'ngcu nhungch(;lYtren n~nInternet. HQc VieDdil cO'g~ngchigiil'l(;linhil'ngyeticftucftnthi€t d~saGchobaitoaDv§:n conhil'ngd~cdi~mthvcte'nhungkhongquaphilct(;lp.Tuynhiendil'li<$uv§:n duQcgiil'nguyennhuhi<$nco, kich thuocdil'li<$ukha IOn:Dil'li<$utrongbang chi ti€t ddnd~thangla 1,533,710m§:utin. Datamining,Clfth~la dmki€m caclu~tk€t hQpla mQtchilcDangcuah<$ thO'ng.Vi duQCtichhQpVaGh<$thO'ngDenchilcDangnayduQCth~hi<$nmQt cachd~hi~uchonguoidungcuO'inh~mt(;losl/ti<$nlQivad~dangchohQkhi sadlfng. Phftndataminingd~dmcaclu~tk€t hQpsedl/atrendil'li<$uchinhcuaddn d~thangvasanphffmduQcd~thang.Cacdil'li<$uClfth~duQcminhho(;lduoi OrderlD ,,' !to,:;',, '4,,' ,CuBtomerNam-..!,~,!f 36778 RAX ERECTING SERVICE INC 36779 RILEY CONSTRUCTION CO INC 36780 RIDGEVVORTH ROOFING CO, 367.81: RAMCORP INC 36782 RAMOS CONSTRUCTION INC, 36783, RADISSON CHGO HOTEL 36784'RIVARD MASONRY INC 36785 RANDOU:'HVvELLS E3LDG,CORP 36786' ROEPENACK & KUVVIK 36787'RPI GENERAL CONSTRUCTION 36788- ,RAMCO GENERAL CQI'JTR INC 36789 RA YMOND & SONS I"I'J,<::",, 36790 S & M CONST, 36791-ALFRED H SANDERMASN CONT 36792, ,JOE SANTUCCI SEVVR CONT 36793 SCHULLO PLUMBING INC 36794 SPANCRETE OF ILL INC 36795 DON STOL TZNER MASON CONST 36796, STOL TZNER PLBG & SEVVER 36797.VVM,J SIEVERT INC 36798, STAMAK CORP 36799: OTTO SCHULTZ 36800' SCHAEFGES BROS INC 3680 SILVESTRI PAVING CO",...... 3680 VV R SEEGREN GENE,RAL CNSTR 36803 SHEFFIELD TANK CO, 4/ 4/5/99 4/5/99 4/5/99 4/5/99 4/5/99 4/5/99 4/5/99 4/5/99, 4/5/99, 4/5/99 4(?!s.,s. 4/?(s.s.. , 4/5/9s., "SentDate::'Y.., 4/6/99 4/6/99 4/6/99 4/6/99 4/6/99 4/6/99 4/6/99 4/6/99 , ,4/6/99 4/6/99 4/6/9,9 4/6/99 4/6/99 4/6/99 4/6/99 4/6/99 4/6/99 4/6/99 4/6/99 4/6/9,9 4/6/99 4/6/99 4/6/99 , 4/6/99 4/6/99 4/6/99 0001','CreatedDateMI#%V 4/5/99' 4/5/99 4/5/99, 4/5/99 4/5/99, 4/5/99 4/5/99,: 4/5/99: 4/5/99 4/5/ 4/ MQtsf{mftutin cuado'nd:}thang Trang57 Minhho~:Xftydt!ngcaemodunkMmpM lui)tke'th(/pchoh~th6ngdi;ithangtrenm~ng Internet 1 2 3 Mc)tsf{m~utin cuachititt mc)tddnd~thang Tli du li~ug6cnayd~thlfchi~nc6ngvi~cdatamining,chuangtrlnhse chuy~nd6iquad~ng0,1thanhbangcod~ngsau.d dayVIchuangtrInhchi quailtamd€n nhungsanphgmcodQh6trQIOnhanhayb~ng0.5liennhung sanphgmcodQh6trQnhohansebi lo~ibokhoibang. Trang58 Minhho~:Xaydt;ingcaemodunkhamphalu~tke'thQpchoM th6ngd~thimgtrenm~ng Internet 1 1 1 0 1 1 1 1 1 1 0 0 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 Dii'li~udu'Q'chuy~nd6iquad~ng0,1 Ap dvngly thuy€tafftrlnhbayatren,mahlnhdataminingttongtruongh<;Jp nayau<;Jccv th~hoanhusau: . R lat~pcacsanphgm. . T la cacdona~t hangau<;Jca~idi~nbaichIdanh. . L la t~pta'tcacact~pconcacsanphgm. . £)i~uki~nq au<;Jcainhnghlanhusau:q(r,X) =truen€u s6cacsanphgm Trang59 Minh ho~:Xay dt!ngcaemodunkham'phalu~Hke'thl!pchoh~th6ngd~thangtren m~ng Internet cuaX duQ~d~thangvdi s6 luQngcacdond~thangIOnhonhaybAng minsup*(tdngso'dondcJ.thimg)/lOO. . Quanh~::;la quailh~c Ke'tquacu6iclingla tlmra duQccaclu~ltke'thQpco d~ngX=>Bvoi X la mQtt~phQpcacsanphftmnaGdova B la mQtsanphftmnaGdo. Vi dl,llu~t234-235->240thlX={234,235},B ={240}va lu~tduQcdi~ngiai thanh: Ne'uale sanphdmco chi danh234(Layersalad)va 235(Macaroni cheese)dlic;fcd(ithang Thzsanphdmcochidanh240(All cheeseappetizertray) cLingdlic;fcd(it hang. VII.2. K[EN TRUe PHAN MEM Ph~ndataminingduQcthie'tke'chiaralamb6nlOpriengbi~t: - Lop giaodi~n(Presentationtier)chuacactrangwebth§ hi~ngiaodi~n cuah~th6ng. - Lop xli 19nghi~pVl,lnghienggiaodi~n(UI-centric)lu'utru thanhph~n DMUICentric chua d6i tuQngUIMining baa g6m cac ham phl,lcVl,lcho giaodi~nhaytuongtacvoi lOpData-centricnhull!a chQncac thams6, gQi cachamtinhtOaD. Trang 60 Minh ho~:Xay dlfngcaemodunkhampM lu~itke'thqpchoM th6ngd~thangtren m~ng Internet Lop xli'ly nghi~pVl;lnghiengda li~u(Data-centric)lu'utra thanhph~n DMDataCentrichliad6i tu<;1ngDMDataCentricbaag6mcachamxli'ly da li~uphl;lCVl;lcha lOp UI-centric va d6i tU<;1ngDMLibDataCentric baa g6mcachamtlmkie'mlu~tke'th<;1ptheagiii thu~tng~unhienhayLevel- WIse. - Lop phl;lCVl;lcacxli'19da li~u:Chlia cd so da li~u. lAp giaodi~n Liip Random,Levelwde) --- ca:c trangWeb ------ DMDataCentrlc (UIMining) (DataMining) Sdd6kit n true Vi~c tri~nkhai ling dl;lngVaa thlfcte'dU<;1chQcvieDthlfchi~ntren 3 server theadungma hlnhdad~nghi: Trang61 Minh ho~:Xfiy dtfngdie modunkhampha]u~Hke'thQ'pehoM th6ngd~thflllg tren m~ng Internet May chu Web (Web server): duqc dli d~tInternet InformationServer (lIS), Microsoft TransactionServer(MTS) va lOpxii'ly nghi~pVl,lnghieng giaodi~n,tilc la chilathanhph§nDMUICentric. - May chu ling dl,lng (Application Server): duqc cai d~t Microsoft TransactionServer(MTS) va lOpxii' ly nghi~pVl,lnghiengdli li~u,tilc la chilathanhph§nDMDataCentricvaDMLibDataCentric. May chu dli li~u(DatabaseServer):duqccai d~tSQL server6.5,chila cd sa dli li~ucuah~th6ng. / Liy thongtin thOngquatrlnhduy~tWeb / Mayt!,!m Maychu Web TCPilP TCP/IP Maychu lfngdlfng Maychu duli~u TCPilP Thanhp~n IDMDmCentric, !DMLibD.:naCen C;icthanhp~nc6t~u9Cl tjn~tmy:tuft11)':ta Trang 62 Minh ho~:Xay dtfngeaemodunkhamphalu~tke'thQ'pehoM thO'ngd~th~mgtren m~ng Internet VII.3.CO sciDOLltU Cacbang(table)cuah~th6ngd~thangco lien quaild€n phgnminingdata baag6mtb_Order,tb_OrderDetailvatb_Productvoi thongtinchiti€t nhu sau: 1. tb_Order M1,lcdich: Luu thongtin t6ngquatcuaddnc1~thang. Tentru'ong OrderID CustomerID CreatedDate SentDate Int(4) Khoachinh Chidanhcuaddnd~thang (PrimaryKey) Int(4) Khoa ngoai Chi danhcuakhachhang (foreignkey) Datetime(8) Khongr6ng Ngaycuaadnd~thang Datetime(8) Ngay thljc hi~n c1dna~t hang 2. tb_OrderDetail M1,lc1ich:Luuthongtinchiti€t cuaadnd~thang Tentru'ong Trang63 Minh ho~:Xay dlfngcaemodunkhampM lu~tke'thQ'pehoh~th6ngd~thangtren m~ng Internet OrderDetailID IInt(4) Khoachinh ChI danhcua chi ti€t don (primaryKey) d~thang OrderID Int(4) Khoangoai ChIdanhcuadond~thang (foreignkey) ProductID Int(4) Khoangoai ChI danhcua san ph~m du'<;1cd~thang(foreignkey) Quantity Smallint(2) I>0 S61u'<;1ngsanph~m Price Real(4) >0 DongiacuamQtdonvi san ph~m 3.tb_Product M\lCdich:Lu'uthongtincuasanph§m Tentrliong Ki~ll trIton ProductID Int(4) Khoa ChIdanhcuasanph~m chinh (Primary Key) CategoryID Int(4) Khoa ChIdanhcualo~isanph~m Trang64 Minh hoq:Xftydlfngcaemodunkhamphalu~itke'thQ'pchoh~th6ngd~thangtren mqng Internet ngoai (foreign key) ProductNaIDe Varchar(50) Khong -ITensanphgm rang Price Real(4) >0 DongiamQtdonvi sanphgm Selected Tinyint(l) ChI dinh san phgm nay co thamgiatrongquatrlnhmining khong. =0:Khong =1:Co Ngoai cac bang chua dITli~u cho h~th6ngd;:ithang,d€ phl;!cVl;!cho vi~c mining data, hQc vien con thie'tke' them mQt so' bangkhac baa g6m Confidence,Mining, Frequent,MiningConfigva MiningStatusvoi thong tinchitie'tnhu'sau: 4.Confidence Ml;!cdich:Bang lu'uke'tquaminingdu'oid,;mgx-> B Trang65 Minh ho~:Xay dlfngcaemodunkMm pM lu:;Uke'thQ'pchoM th6ngd~th~mgtren m~ng Internet x Varchar(255) T~pX B Varchar(10) T~pB 5.Mining Ml;lcdfch:Banglu'uke'tquacuabuC1ctinhtoanc1p1trongvi~cdatamining. 'fentruong Kiiu buQctru'ong OrderID Int(4) ChIdanhcuaddnd~thang Selected Tinyint(l) ChI dinhddnd~thangnayco tham gia trong vi~c mInIng datahaykh6ng. =0:kh6ngthamgia =1:cothamgia F Tinyint(l) i=1tom vC1imla sf)sanph~m tlm dU<;1Csau buC1chq.yd§u tien. Trang66 Minhho~:XaydtfngcaemodunkMm pM lu~tke'thQ'pchoM th6ngd~thfmgtrenm~ng Internet Ni la Chi danhcuasanphim UrndU9C F=l. sanphim naydU9C d~thangtrongdond~thang =0:Khongdu9Cd~thang 6.Frequent M\lc dich:Bangdungd~hill ke'tquatrunggiantu bangMiningra ke'tqua cu6icunglu'utrongbangConfidence ID Int(4) Khoa Chi danhcuam~utin chinh (primary key) Fj, j=l toim Int(4) Cac san phim tlm du9C,j 100 thli tv cac sanphim trongbang Mining Setlevel Tinyint(1) Ca'p(s6sanphim cuatZiph9P) Trang67 Minhho~:Xaydtfngcaemodunkhamphalu~tke'thQ'pchoM th6ngd~thangtrenm~ng Internet 7.MiningConfig MlfCdich:Bangdungd~lliucacThamsf{phlfCVlfchovi~cminingdata Tentniong Support Float(8) >=0.5 DQ h6 trCJ Confidence Float(8) >=0.5 DQtinc~y OrderStartDateDatetime(8) Ngay bilt dftukhi co gioi h~n vi~ctinhtoancacdond~thang OrderEndDate Datetime(8) Ngay ke'tthuckhi co gioi h~n Trang68 Support Tinyint(l) DQ h6 trCJtuongling voi t p tImduCJc Type Char Bie'ndungtrongvic tinhtoan theogiaithut Randomized - Type=O:t pC - Type=l:tp bucuatp C - Type=2:tp X - Type=3:tp bli cuatp X Minh ho~:Xay dtfngcaemodunkMm pM lu~tke'thQ'pehoh~th6ngd~thangtren m~ng Internet vi~ctinhtoancaedond~th~lllg Kl Tinyint(l) >0 Thamsf{kl sadl,mgtronggiai thu~t Randomized K2 Tinyint(l) >0 Thamsf{k2 sadl;lngtronggiai thu~tRandomized BitmapCreated Datetime(8) NgaygiGt~obangMining OrderCount Int(4) sf{dond~thangthamgiaVaG quatrlnhmining ProductCount Int(4) sf{sanph5mthamgiaVaGqua trlnhmining Method Tinyint(l) Lo~i giai thu~tsa dvng cho vi~cdatamining. =0:Level-wise =1:Randomized 8.MiningStatus Ml;lCdlch:Lu'utr~ngthaicuaquatrlnhmining.N€u congu'oidangthlfchi~n quatrlnhminingthlh~thf{ngsechIchoxemk€t quacualffnminingtru'ocdo. Trang69 Minh ho~:Xftydl!ngcaemodunkhampM lu~tke'thqpchoh~th6ngd~thangtren m~ng Internet DateTimeCreated Datetime(8) NgaygiOcuaIfinminingcu6i cling Status Tinyint(1) Tr'.lngth,iicuaquatrlnhtht;1'c hi~nmining =0:Da tht;1'chi~nxong =1: Dangtht;1'chi~n Trang70 MinhbOIl:Xiy dtfngcacm&dunImamphal~t k~thcjpchoh~th6ng~t hangtr~nmllng Internet VII.4. GIAO DItN Giaodic$ncuaphcinDataminingbaog6mcaemanhlnhsau: 1.Man hlnhchinhcuaDataMining(DataMiningpage) M\le weh: Cho phepngrtCJidungxemke'tqua,tht!chic$nvic$cmininghay thayddicaethams6ph~c~ chovic$cmining. Caethaome: - Nha'nnutView Resultd~xemke'tquaminingcualcintht!chic$nmining trrtoc. - Nha'nnutMined~tht!chic$nvic$cminingdiYlic$u. - Nha'nnutOptionsd~xemhaythayddi caeThams6ph~cv~chovic$c miningdiYlic$u. - To change the options for data mining,click Options - To view result of the last mining,click View result - To start miningdata, click Mine Trang 71 Minhho,: Xiy dtfngcacmodunkhamphalu4tk~thcjpchoh~th6ngtift hangtr~nm,ng Internet 2.Man hInhxemke'tquaDataMining(MiningResultpage) M\le dieh: Chophepng1toidungxemke'tquacuah~nth1!chi~nminingtntdc d6.Ke'tquad1t<;fcth~hi<$ntrenmanhinhd1tdid~ngbangbaog6mhaicQtX vaB tUdng1i'ngvdilu~tX->B. Caethaome: Nha"nnutOptionsd~xemhay thayd6i cacThams6phq.cvq.chovi<$c miningdfi li<$u. Nha"nvao lu~ttUdng1i'ngtrongcQtX d~xemdi~ngiai cualu~t. - Tochangetheoptionsfordatamining,clickOptions Mining Result12/Ui/1999 22:05:47, {Running time [0:38]) - Toviewthee.planationofa rule,pleaseclickthecorrespondingiteminX column ¥-= .'iH~1 Trang 72 Minh ho:;t:Xay dQ'ngcaemGdunkhamphalu~Hke'thqpchoM th6ngd~thangtren m:;tng Internet 3. Man hinh xem di~n ghH lu~itke't hQp tim duQc (Rule Explanationpage) M\lC dich: Chaphepnguoidungxemdi~ngiai cualu~ttlmdu<;1crheangii' nghlathongthuong. Vi dl;llu~t234-235->240dU<;1cdi~ngiaithanh: Niu cacsanphdm"Layersalad"va"Macaronicheese"dU:(jcdijthang Thlsanphdm"Allcheeseappetizertray" dingdU:(jcdijthang Cac thaotac: Nh§nnUtClosed~trdv~manhlnhxemke'tquamining(MiningResult) truckdo. IF Layer salad Macaronicheese ARE ORDERED THEN All cheese appetizertray IS ORDERED Q.;>M 1 Trang73 Minhho~:XaydQ'ngcaemodunkMmpM lu~tke'thc;tpchoM th6ngd:}thangtrenm~ng Internet 4.Man hinhc~pnh~tcaethams6chomining(Optionspage) M\lc dich: Chophepngu'oidungxemhayc~pnh~tcacthamsf{phl,lcVl,lcho vi~cminingdli li~utheo2giaithu~t:ngftunhienvaLevel-wise. Cac thaotac: Nha'nnutLimit Orders d~gioi h~nph~mvi cacdond<)thangthamgia trongquatrlnhminingdata. Nha'nnutLimit Productsd~gioih~ndanhsachcaesanphffmthamgia trongquatrlnhminingdata. Nha'nnutSubmitd~lu'ucaeThamsf{vaoh~thf{ngva quaytrdl~iman hlnhchinhcuaDatamining(DataMiningpage). Nha'nnUtCancel d~quayv~manhlnhchinhcua Datamining(Data Miningpage)makh6nglu'ucacsl'tad6iv~thamsf{vaoh~th6ng. - To limit the numberof orders, click Limit Orders. - To limit the numberof products, click Limit Products, - After finishingediting, click Submit. - To leave this page without makingchanges, click Cancel. Algorithm Support Confidence Number of rows r Randomize ~ ~ [ [ [ [ r level-wise [O.S-l] [O.S-l] .Ym!!.Cil~f Ym3!",(~fNumberof columns K1 K2 S\.I~f ~iI<!, '1 Trang 74 Minh ho~:Xay d1!ngcaemodunkMm pM Iu~tke'th<,1pchoM th6ngd~thangtren m~ng Internet 5.Man hinhgioih~ncaedond~thangthamgiaVaGvi~cmining (LimitOrderspage) MQ.cdich: Chophepngu'aidunggidih(;lns6dond~thangthamgiaVaGqua trinhminingdli li~u. Ngu'aidungco th~chQncacdond~thangn~mtrong me)tkhoangthaigianba'tky tungaynaGtoingaynaGhaychQntoanbe)ta'tca cacdond~thang. Cac thaotac: - Nha'nnutSubmit d~lu'ucacthams6 v~vi~cgioi h(;lns6dond~thangVaG h~th6ngva quaytrdl(;limanhinhc~pnh~tcacthams6phl,lcVl,lchodon d~thang(Optionspage). - Nha'nnutCanceld~quayv~manhinhc~pnh~tcacthams6phl,lcVl,lcho dond~thang(Optionspage)tru'ocdomakh6nglu'ucacthams6 v~vi~c gioih(;lndond~thangVaGh~th6ng. - After finishingentering, click Submit. - To leave this page without makingchanges/adding,click Cancel r Allorders r Onlyordersfrom to [ ~t :,. ~"...<I Trang75 Minh ho:;1:Xiiy dQ'ngcaeme>dunkh3mph3lu~tke'th~pchoh~th6ngdijt hangtren m:;1ng Internet 6.Man hinhgioih~ncacsanphftm(LimitProductspage) Ml)c dich: Chophepngu'oidunggiOih<:lndanhsachcacsanphftmthamgia vaoquatdnhminingdil li~u. Ngu'oidungco th~chnta'tca cacsanphftm hayChImQtsf)sanphftmnaodothai. Cac thaotac: Nha'nvaocaccheck-boxtu'onglingvOitungsanphftmdu'<;jcli~tke tren manhlnhd~chQnlbochQnsanphftm. Nha'nnutSelectAll d~chQnta'tcacacsanphftm. Nha'nnutDeselectAll d~bochQnta'tcacacsanphftm. Nha'nnutSubmitd~h1ucacthamsf)v~vi~cchnsanphftmthamgia trongquatrlnhminingdil li~uvaoh~thf)ngva quaytrdl<:limanhlnhc~p nh~tcacthamsf)phl;}cVl;}chodond~thang(Optionspage). Nha'nnutCancel d~quayv~manhlnhc~pnh~tcac thamsf)phl;}cVl;}cho dond~t hang(Optionspage)tntocd6makhanglu'ucacthamsf)v~vi~c gioi h<:lncac sanphftmthamgia vao quatrlnhminingdil li~uvao h~ - ClickthecheckboxinSelectcolumnto select/deselectproduct. - To selectallproducts,clickSelectAll. - To deselectallproducts,clickDeselectAll. - After finishingentering,click submit. - To leave this page withoutmakingchanges/adding, click Cancel. . roduct NamS',ViWJ#P#PN,;,yti/iWPPPitii,'iiNtWNPY'diY:i\'if(tiii'";\'YN'>"";"""",Select'?!:.;}"': r %~~4 ~N¥I $tI.~. J ~'JJ Minh ho~:X~ydt!ngcaemodunkhampM lu~tke'thc;tpchoM th6ngd~th~lDgtren m~ng Internet VII.5.CAC THANH PHAN (COMPONENT)vA DOl Tu'(1NG (OBJECTS) Ph~nnaymatacacthanhph~nvatungd6itU<;1ngtrongm6ithanhph~nphl;!c Vl;!chovit$cDatamining. ", ' 1. Cac thanhphan C6 3 thanhph~nphl;!cVl;!chovit$cDatamining. Chungdu<;1cmata trong bangsail: Tenthanhph~n Mot3 DMLibDataCentric Phl,!c VI,!rho vit%cmining dft lit%utheo hai giai thu~t Level-Wiseva ng~unhien(Tinh roanW bangMining m >Frequent--->Confidence). DMDataCentric Phl,!cVI,!rho vit%ch!a chncac Tham s6 v~phia dft lit%u. Tinhtoan/t(,to/c~pnh~tbangMiningla bangchuacac sanphimcap1. DMUICentric Phl,!c VI,!rho vit%eIlfa ehQncaeThams6 vephiagiao dit%n. 2. D6i tu'qng(Objects) CacdoitU<;1ngcuatungthanhph~nphl;!cVl;!chovit$cDataminingdu<;1cmata trongbangsail: Trang77 Minh ho~:Xay dtfngeaemodunkMm pM lu~tke'thc;lpehoh~th6ngd~thangtren m~ng Internet DMLibDataCentric I LevelWise Public Tinh roan mining theo giai thu~t Level-wise. Random Public Tinh roanminingtheogiai thu~tngIlu nhien. DMDataCentric DataMining Public Tho'vi~ncacham,thuQctinhcuavi~c miningv~ho'ongdITli~u. DMUICentric UIMining Public Tho'vi~ncacham,thuQctinhcuavi~c miningv~ho'ongiaodi~n. 3. Thu()ctlnh (Attribute) Chi co d6i tu<;1ngUIMining cuathanhphftnDMUICentricla co thuQctinh. ChungdU<;1cmotatrongbangsail: BitmapCreatedVariant Public Public Ngay giGt(,lObangMining Columns Integer Public s6 cQt(s6sanphftm)thamgia vaovi~cmining. Trang78 Minhho~:Xayd1;ingcaemodunkhamphalu~tke'thQ'pchoM th6ngdiJ.thangtrenm~ng Internet Trang79 Confidence Variant Public Public DQtincy (Confidence)t6i thi6u (minconf) IsDirty Boolean Public Public Bt/tiit coco sliad6idli lit%u. IsNew Boolean Public Public Bt/tt coco themdli lit%u. IsValid Boolean Public Public B t/d t co co dli lit%uco hcJpIt% kh6ng. .. Kl Boolean Public Public Tham sf) Kl trong giiii thut ng unhien. K2 Boolean Public Public Tham sf) K2 trong giiii thut ng unhien. MiningMethod Byte Public Public Giiii thut chQnchovit%cmining (1:Level-wIse,2:ngu nhien) OrderCount Long Public Public so'dondt hangdu'cJchQn. OrderEndDate Variant Public Public Ngay cu6i trong phm vi h,fa chQnngayd6gioi hn so'dondt hang tham gia vao qua trlnh mInIng. OrderStartDate Variant Public Public Ngay dftu trong phm vi h,fa chQnngayd6gioi hn so'dondt hang tham gia vao qua trlnh mInIng. Minh ho~:Xay dl}'ngcaemodunkhamphalu~tke'thQ'pchoM th6ngd~th~mgtren m~ng Internet Product Long Public La'y Chi danh cua san ph~m tu'onglingtli trongmangcacsan ph~mdu'QcchQn. ProductCount Integer Public S6 sanph~mdu'QchQnchovi~c miningdli li~u. Support Variant Public Public DQ h6 trQ (support)t6i thit€u (minsup) 4. Ham(Method) Cac ham cua tung d6i tuQngcua tungthanhphgn ph1,1cV1,1cho vi~c Data miningduQcmatatrongcacbangsail: a) DO'itu'Q'ngLevelWise: ThuQcthanhphgnDMLibDataCentric Tenham LevelWiseCal Public Ham tinh roan thea giai thu?tLevel- WIse Thams6: alngProductListO(Long):Mangcac sanph~mdn tinhroan. lngRow(Long):S6 dong(s6dond~t Trang80 Minh ho~:Xay dtfngcaemodunkhamphaIu~tke'thQ'pchoM th6ngd~thangtren m~ng Internet hang). - sngSupportMax(Single):DQh6trQ. - sngConfidenceMax(Single):DQ tin c~y. - IngErr (Long): Ma: 16itra v~(=0: Kh6ngco 16i,khac0: Co 16i) ConfidenceCal Private Ham tim cac t6 h<;1pva tinhtoaDdQh6 trQvadQtinc~ytu'onglingvac~pnMt vanbangFREQUENT) Thamso: - sngConfidenceMax(Single):DQ tin c~y - IngErr (Long) : Ma: 16i tra v~(=O: kh6ngco16i,khac0: co16i) CreateCandidate Private Ham tinhtoaDt~pcac ling VieDva du'a vanbangCONFIDENCE dtjatrenbang FREQUENT. Thamso: - intColums(Integer):SocQt - intSupportMax(Integer):DQh6trQ - IngErr (Long) : Mil 16i tra v~(=O: Trang81 Minh ho~:Xay d1!ngcaemGdunkhampM lu~itke'thqpchoM th6ngd~thangtren m~ng Internet khongco16i,khac0:co16i) b) D6i tu'Q'ngRandom: ThuQcthanhph~nDMLibDataCentric RandomCal Public Ham tinh roanrheagiai thu~tng~u Tenham Ki~utrj I Ph~mvilMo fa trav~ nhien. Thams6: - alngProductListO(Long):Mangsan phftmdin tinhroan - IngRow(Long):S6dong - intKl (Integer):Thams6Kl - intK2 (Integer):Thams6K2 - sngSupportMax(Single):Dc)h6trQ - sngConfidenceMax(Single):Dc)tin c~y - IngErr (Long) : Mti 16i tra v~(=O: khongco 16i,khac0: co16i) ArandomMFS Private Ham tinh roan rhea giai thu~t A- Trang82 Minh hm~:Xay dtfngcaemodunkhampM lu::itke'thQ'pchoh~th6ngd~thangtren m~ng Internet Random-MFS(S) Thams6: - abytSO(Byte):Mang t~pS - intK (Integer):Thams6K - intColums(Integer): S6 cQt - mtSupportMax(Integer):DQh6trc;l - lngTotalRow (Long): S6 dong (s6 aonaat hang) - lngErr (Long) : Mil l6i tra v~(=O: kh6ngco loi, khac0:co loi) Complement Private TIm t~p hc;lPbil abytD cua t~p hc;lP abytC Thams6: - abytCO(Byte):T~phc;lP(mang)dn lay t~pbil - abytDO(Byte):T~pbil (ke'tquacua ham)cuat~pabytC - intColums (Integer): S6 cQt (S6 ph~ntudn lay t~pbil) GenerateRules Private Ham tinh loan alia ra cac lu~tke'thc;lP Trang83 Minh ho~:Xay dt!ngcaemodunkMm pM lu~itke'th(lpchoM th6ngd~thangtren m~ng Internet va htu vao bang CONFIDENCE dt!a teenbangFREQUENT. Thamso: - intColums(Integer):SocQt - sngConfidenceMax(Single):DQ tin C?y - IngErr (Long):Ma 16i teav~(=O: kh6ngco16i,khac0:co16i) MaximalFrequent Private TIm t?P toi d<;li Thamso: - intColums(Integer):SocQt - IngErr (Long):Ma 16i teav~(=O: kh6ngco16i,khac0:co16i) Permute Private Hoanvi t?P (mang)abytSthanhabytRS Thamso: - abytSO (Byte): T?p (mang) dn hoanvi - abytRSO (Byte): T?p (mang) k€t <. qua AssignArrayBoolean Private Gan gia tri cua mang ablnArr2 rho Trang84 Minh ho~:Xay dgngcaemodunkMm pM lu~tke'thQ'pchoh~th6ngd~thangtren m~ng Internet ablnArrl Thams6: - ablnArrl (Boolean):Mang dn gall gia tr!. - ablnArr2(Boolean):Mangdn la'y gia tr!. - n (Byte): S6 ph<1ntti'cua mangdn gall gia tr!. AssignArrayByte Private Gan gia tr! cua mang abytArr2 cho abytArr1 Thams6: - abytArrl (Byte):Mang dn gall gia tr!. - abytArr2(Byte):Mangdn la'ygia tr!. - n (Byte): S6 ph<1ntti'cua miingdn gallgia trio SortArray Private S~pxe'ptheothli tt! tangd<1ncua mQt mango Tham 56: - abytS(Byte):Mangdn s~pxe'pthu Trang8S Minh h()~:Xay dlfngcaemodunkhamphaIu~tke'thQ'pehoM tho'ngd~thimgtren m~ng Internet c) D6i hiQ'ngDataMining ThuQc thanh ph§n DMDataCentric Tenham Ki~utri I Ph~mvi tnl v~ CountOrder Long Public ExtractData Public GetConfigInfo Public ttf - intColums(Byte):So cQt(phftntti' cuamang)dn s5pxep. Dernsodond~thangduQcchQn. Tharnso: strStart(String(10»: Ngaydftu. strEnd(String(10»: Ngaycuoi. Tinh roandQhe;trQcua tungsanphirn va t'fObangMINING Tharnso: p_sngSupport(Single):DQhe;trQ vrFinish(Variant):Tra v~ngay& giot'fobangMINING. uiy cacthalliso cuah~thongtubang MININGCONFIG Trang86 Minh hm~:Xay dtfngcaemodunkMm pM lu~tke'thQ'pehoM thongd~thangtren m~ng Internet GetMiningProduct GetSelectedProduct MiningStatus Integer SaveConfigInfo Public Public Public Public Thams6: - arrResultO(Variant):Mang trav~gia trtcuacacthallis6. U1Ydanhsachcacsanphim Thams6: - arrResultO(Long):Tra v~mangcac sanphim. uiy danh sach cac san ph:1m duQc chQn. Thams6: - arrResultO(Long): Tra v~mangsan phim auc;!chQn. UIY tInt tr<:lngcua qua trlnhminingdli lit%u.TInh tr<:lngnay auc;!cluu d truong StatustrongbangMININGST ATUS. Lu'ucacthallis6cuaht%th6ngdungcho vit%c dli lit%u bangvaamInIng MININGCONFIG. Thams6: - sngSupport(Single):DQh6trc;! Trang87 Minh ho~:Xay dlfngcaemodunkhamphalu~tke'thQ'pchoh~th6ngd~thangtren m~ng Internet SetMaskProduct Public SetOrderSelected Public - sngConfidence(Single):DQtinc~y - strStart(String(10)): Ngay b~td~u dungtrongvi~cgidi h<;tns6 dond~t hangthamgiatrongvi~cmining. - strEnd(String(10)):Ngay cu6idung trongvi~cgidi h<;tns6 dond~thang thamgiatrongvi~cmillIng. - bytKl (Byte):Thams6Kl - bytK2(Byte):Thams6K2 - bytMethod(Byte): Giai thu~tdu'Qc ch<;m(1:Level-wise,2:ng~unhien) C~pnh~ttlnhtr<;tngIt/achQncacsan phftm (tru'ongSelectedtrong bang tb_Product) Thams6: - arrProductID(Variant): Mang san phftmdn c~pnMt tlnhtr<;tng. - arrMiningPro (Variant): Mang san phftmtrav~. C~pnh~ttint tr<;tngIt/a ch<;mcac san phftm (tru'ong Selected trong bang Trang88 Minh ho~:Xiiy dt!ngcaefiG dunkhamph:!lu:)tke'thQ'pehoh~thO'ngd~thilllgtren m~ng Internet MINING) Thamso: - p_strStartDate(String (10)): Ngay bittogudungtrongvi~cgioi hi;inso odn o~thang thamgia trongvi~c mInIng. - p_strEndDate(String (10)): Ngay cuoidungtrongvi~cgioi hi;insoodn o~t hang tham gia trong vi~c mInIng. d) D6i tu'Q'ngUIMining ThuQcthanhphgnDMUICentric Tenham Ki~utri I Ph~mvi IMota trav~ . ApplyEdit Public Tien hanhvi~cC?Pnh?t. BeginEdit Public B?t che09chophepC?Pnh?t. CancelEdit Public Tilt che09chophepC?Pnh?p ExtractData Public Thlfc hi~n vi~c gQi t<;iobang MINING Trang89 Minh ho~:Xftydt!ngcaemodunkhamphrilu~Hke'thQ'pehoh~th6ngd~thangtrenm~ng Internet GetProductList Public LoadConfig Public MininglnProcess PublicBoolean ProcessMining Public SaveConfig Public SetMaskOrder Public tablerheacacthongsocuah~thong. Lay danhsachcac sanph~mthamgia trongquatrlnhminingdli li~u Thamso: - arrProduct(Variant):Mang sanph~m trave Lay cacthongso cuah~thong(du'9c lu'utrongbangMININGCONFIG) Ki~mtraxemquatrlnhminingdli li~u co dangthlfc hi~nkhong (Lu'u trong tru'ong cua bangStatus MININGST ATUS) Thlfc hi~nvi~cgQi thu t1,lCthlfchi~n miningrheagiai thu~tLevel-wisehay ngaunhien. . Lu'utrli cacthongso cuavi~cmInIng diYli~uVaGbangMININGCONFIG. C~pnh~ttlnh tr<;1ngllfa chQncac don dZithangph1,lcV1,lcho vi~cminingdli li~u. Trang90 Minh ho~:Xfiydt!ngcaemodunkhampM lu~Hke'thQpchoh~th6ngd~thangtren m~ng Internet SetMaskProduct Public C~pnh~ttlnh tr"mgIlfa chQncac san ph§'mph\lCV\lchovit$cminingdii'lit$u. Trang91