yy yfai H^i thto ICT.rda'06 Proceedings of ICT.rda'06 Hanoi May 20-21.2006 TICH HQp PHlTOfNG PHAP VAO ]Vl6 HINH CO s d DOI TirOfNG XAC SUAT ]Vl6f Integration of Methods into a Fuzzy and Probabilistic Object Base IVIodel Nguyen Hoa, Cao Hoang Tru Tom tat Bdi bdo ndy gidi thieu mgt md hinh ca sd ddi tugng xdc sudt md (fuzzy and probabilistic object base - FPOB) cd tich hgp phuang phdp (method) vdo Irong cdc lap ddi lugng Di Idm nhu vgy, chiing tdi dd coi mdi phuang phdp nhu mgt dnh xg md gid tri cita nd Id mgt phdn bd xdc sudt tren mgt Igp gid tri bieu dien hdnh vi khdng ehdc chdn cua cdc ldp ddi lugng Theo dd cdc khdi niim kieu luge dd ca sd doi tugng xdc sudt ma, thi hiin vd phep chgn dugc dinh nghTa di cho phep truy vdn ddi tugng thdng qua viec thyrc thi cdc phucmg phdp Tir khod: Ca sd ddi tugng xdc sudt ma, xdc sudt, tap md, phuang phdp, truy vdn Abstract This paper introduces a fuzzy and probabilistic object base model (FPOB) that integrates methods into classes In order lo realize this, we have considered each method as a mapping whose value is a probability distribution over a set of values, representing uncertain behaviors of object classes Then the notions of types, fuzzy-probabilistic object base schemas, instances, and selection operation are defined for querying objects via execution of methods Keywords: Fuzzy and probabilistic object base, probability, fuzzy set, method, query GIOI THIEU Nhu chiing ta da bilt, md hinh hudng doi tugng truyin thong da chung td cic uu dilm cua nd cic vin dl md hinh hda, thiit kl vi hifn thyc cic hf thong Idn, tir phin mlm cho din CO sd dir lifu Dk bilu diln vi suy luin tren cic thdng tin khdng chic chin va khong r5 rang phd biln thyc tl, nhilu nghien curu da va dang dugc tiln hanh dl md rpng md hinh hudng doi tugng truyin thong bing each ip dyng cic kit qui cua ly thuyet xic suit va ly thuylt tip md Mpt so nghien curu cho phep gia trj thupc tinh cua cac doi tupng li nhirng tip md, nhien mdc dp khdng chic chin dl doi tugng cd cic thupc tinh md nhu viy chua dupe quan tam ([2], [4], [8], [11], [12])^ Mpt so nghien ciiru diing xac suat dl bieu diln tinh khdng chic chin ciia sy cd mit ciia mpt thupc tinh mpt Idp doi tupng va ap dyng ly thuyet xic suit, ly thuyet jgp ^(j ^g jjj^u (jj|„ p^c gi^ trj n^^ ^rong thupc tinh dd ([1], [5]) Mpt so nghien cuu khic quan tam din ciy phin cap cic ldp doi tupng bing each xem xet xac suat dl mot doi tugng thupc ldp cha la thupc vl mpt ldp con, din den khii nifm ciy phan cip vdi xac suat ([3], [9]) Gin diy chiing tdi da md rpng md hinh [9] cho cic thupc tinh ciia doi tupng co thi nhin gia trj md, dan den mpt md hinh co sd ddi tugng xac suit md vdi ciy phin cip xac suit ([6], [7], [10]) Trong bai bio niy, chiing tdi tich hgp cic phuong phip vio md hinh niy, dl cd the bieu dien dupe hanh vi cua cic ldp doi tugng vi truy vin doi tugng thong qua vifc thyc thi cic phuong phip D I cd thi tich hgp cac phuong phip vio cic ldp doi tugng, phin 2, chiing toi md rpng khai nifm kieu bg (tuple type), gj^ trj, gid tn bg xdc sudt md (fuzzy- Ky ylu HQi thto ICT.rda'06 probabilistic hiple value), luge tren tip cic tinh chdt (property) li thudc tinh hoic phuong phap, thay vi tren tip cic.thupc tinh nhu da dupe djnh nghTa [7] Sau cic khai nifm mo hinh se dugc md rpng mpt cich tuong iing dl cho phep truy vin doi tugng thong qua vifc thyc thi cac phuong phip nhu cic phin 3,4, Proceedings of ICT.rda'06 Hanoi May Ngoai chiing ta cung dk y ring m the vira thupc LETTER vi PRIORITY doi tugng thudc Idp LETTER_PRI0RIT^ Ciing nhu md hinh hu tugng truyin thong va cac md rpng md ciia nd, mdi Idp dugc die trung bd thudc tinh ma cac gia trj ciia chiing kilu tuong ling nao Trong FPOB chi khai niem kilu tap hgp dugc n CAC KlfeU VA LUgfC D FPOB thinh kilu tip hgp md va gia trj k Doi vdi FPOB chung tdi sfir dyng cung thinh gia trj tip hgp md, mi khai nii djnh nghTa sy phin cap ldp nhu [7] bp cung dugc md rpng dk cho phep t Hinh chi mpt vi dy vl sy phan cip cac gdi cic phuong phap vao cac ldp doi tuc buu kifn (package) [4], mOt md hinh co cac djnh nghTa d phin tiep theo sd ddi tupng xic suit truyen thong, di dugc Djnh nghia 2.1 Gii six A la mpt thay ddi cho phu hgp vdi md hinh dl nghj, d tinh chit va T li nipt tap cic kiiu d6 package (goi) dugc phin lo^i nhu li letter (atomic type) Cac kilu dugc djnh ngi (thu), box (hpp) hoic tube (ong giiy) hay nhu cich qui nap nhu sau: la priority (goi uu tien) hoic normality (gdi Mpi kilu CO sd T li mdt kil thong thudng) NIU T la mpt kilu, thi {x} la mc tip hgp md (fuzzy set type) cua x gpi la kieu tap hgp (set type) NIU y4i, , Ak la cac tinh chat dc khic A va Xi, ,x* OJ, 03/ 03 07 03/ kilu, thi X = [Ai: x, , A^: x*] 1; unu BOX NOMtAUTY lUBE Fuoun kieu, dugc gpi li kilu bp tren ti tinh chat {Ai Ak} Vdi mpt kil 03V 0T~ 0.7 'IJ3 [Ai: xi, , Ak Xk], chiing ta sir dyn; Lrmi^FRDRnT BomnnuuiT dl bieu thj X, Chiing ta gpi Ai, , cac tinh chat miic cao nhdt (top-lev Hinh 1: Mgt vidvphdn cdp ldp FPOB Trong djnh nghTa nay, khai nifm ki^ Cic ldp nhu viy ciia mpt ldp dugc da dugc md rpng tren tip cic tinh cha lien kit vdi mpt nut d li lo^i triir lin khdng chi bd hep tren tip thupc tinh nhu t (nghTa li mpt dii tupng khdng thi thupc vl [7] Chiing ta cung cd thi coi mdi thudc hai ldp t?ii cung mpt thdi dilm) Trong vi dy li mpt tinh chit khdng cd doi sd niy ldp PACKAGE cd hai nhdm ldp li Vi du 2.1 Trong co sd doi tugng cac {LETTER, BOX, TUBE} va {PRIORITY, buu kifn da gidi thifu d tren, cac tinh cha NORMALITY} thi la origin, destination, time, level m Cic gia trj so [0, 1] tren cic cung noi di, ncri din, thdi gian va mire dp uu lien ket giiia mpt ldp vdi ldp tryc tilp cua vin chuyin ciia mpt gdi va mpt sd tinh c nd bilu diln xac suit cd dieu kifn dl mpt ddi khac length, width, height, area, voiu tupng thupc ldp cha li-thupc ldp ciia nd md ta chilu dai, chilu rpng, chilu cao, c Ching h^n, sy phin cip niy chi ring mpt tich, the tich cua mpt goi v.v Trong dd a doi tugng bit ky ciia PACKAGE cd 70% va volume la cic phuong phap tinh difn t ning thupc vl NORMALITY chi cd vi thi tich ciia mpt gdi Mpt so kilu co 30% ning c6n lai thupc vl PRIORITY T cd the la integer, real va string ^ w y f a HQi thto ICT.rda'06 96 kilu tip md vi kilu bp cd thi li {real}, [origin: string, destination: string, time: real] vi [origin: string, destination: string, time: real, length: {real}, width: {real}, area: {real}] Dinh nghia 2.2 Mdi kieu co bin x e T c6 mpt miln xic djnh dom{x) kit hgp vdi nd Gii trj dugc djnh nghTa mot cich qui nap nhu sau: Vdi mpi kieu eo ban x e T, thi mpi v e dom{x) la mpt gii trj kilu x Vdi mdi x e T mpi tip md tren dom{x) li mpt gii trj kilu {x} NIU A , Ak la cac tinh chit ddi mpt khic trong^4 va vi, , v* la cac gia trj tuong irng ciia cac kilu Xi, , x* thi [A^: Vl, , Ak vJ la mpt gii trj kilu [/i,: Xi.-.Ak.Xk] Chiing ta ciing cd the coi mpt tip cd diln A tren mpt miln U nhu la mpt tip md die bift Af tren f/vdi him vien dugc djnh nghTa bdi V X G U, Aj{x) = nlu x e Ava Aj{x) nlu X € A Tuong ty nhu the, mpi v e U ciing cd thi coi nhu mpt tip md die bift V/ tren U vdi him thinh vien dupe djnh nghTa bdi V A: E U, v/x) = nlu x = vva vjix) = nlu x^v Vi du 2.2 Gii sir about_2 va about_3 la cic nhdn ngon ngir (linguistic label) cua cic s6 md tren : x), la mpt anh x?/dit moi ^ bp ba xic suit md {{Vt, at, fit), , {Vk, Ck,fik))Xuangumg tren x,, , Xk vdi mpt bp ba xic xuit md {V,a,P) tren x Vi du 2.3 Gii su, chiing ta bilt mpt goi se dupe chuyin tir Saigon den Hanoi, nhung khdng chic thdi gian la bao nhieu, nhien, nlu chung ta dam bio ring 20-60% ning nd se dupe chuyin thdi gian 15,16 hoic 17 gid thi chiing ta cd thi bilu diln thdng tin bdi mpt gii trj bp xac suit md [origin: {{Saigon}, u, u), destination: {{Hanoi}, u, u), time: ({15, 16, 17}, 0.6M,1.8M)] Trong uja ham phan bo chuin vi 0.6M vi 1.8M bilu dien cac ham phan bo xic suit a vi him fi cho aix) = 0.6(1/3) = 0.2 vafi{x)= 1.8(1/3) = 0.6, V;ce{15, 16, 17} Bay gid lupc FPOB dupe djnh nghTa dl tich hgp phuong phap vao cac ldp doi tugng nhu sau: Proceedings of ICT.rda'06 Hanoi May Ky yiu HQi thto ICT.rda'06 Djnh nghia 2.4 Mpt Itrgc ^d (schema) co sd doi tupng xic suit md li mpt bp nim S = (C, X, =>, me, (p ), dd: F = { V = V , V | V I G K , , V2GF2 } [a(v),Av)] = [«i(v.),>9i(v,)] filiVl)], VV = V1.V2, V,6 K,, V2G C l i mdt tip hiru h^n cac Idp (do li cac ldp dupe ket hgp vdi FPOB) return (F, a, ^ Mdt cich tucmg ty phuong pha X la inh xa tir C din tip eac kilu bp x{c) cic ldp BOX va BOX_NORMAI (cho bilt tinh chit v i kilu dir lifu ciia cd thi dugc djnh nghTa nhu sau: moi Idp) volume((F,, a,, fit): {real}, {V-, => li mpt quan hf hai ngdi tren C cho (C, =>) li mpt thj cd hudng {real}, {Vy, a^, fiy): {real}): {{V khdng CO chu trinh Mdi mdt node cua {real}) (C, =>) la mpt ldp C, mpt cfinh ct K = { V = V I V V | V I G F I , V2GF'2, => C2 nghTa la C| l i ldp trvrc tiep ciia [a(v).Av)] = ([«i(v.),y?i(v,)]) vdi mpt sd hiru ti khoing [0,1] cho 'LdePp{d, c) < 1, Vc e C,yP e meic) Ky hifu Ct =>* Ck vdi ^ > nlu cd mpt dudng di ci=> C2 => => Ck (C, =>) Die bift c =>* c, V c G C Vi du 2.4 Mpt lupc FPOB cho co so doi tupng cic gdi biru kifn d i neu tren cd thi dupe djnh nghTa nhu sau: C = { PACKAGE, LETTER, BOX, TUBE, PRIORFTY, NORMALITY, LETTER_PRIORITY, BOX_NORMALITY } /?2(V2)])®[a3(V3),/?3(V3)],Vv = V , e F , , V2GK2,V3€K3 return (P', a, ^ Bang : Phep gan kieu x c PACKAGE [origin: string, destination: • time: real] LETTER [origin: string, destination: s length: {real}, width: {real} real, area: {real}] BOX [origin: string, destination: si length: {real}, width: {real}, height {real}, time: real, voi {real}] TUBE [origin: string, destination: sti radius: {real}, height: {real}, time: real, volume: {real}] PRlORnY [origin: string, destination: sti level: integer, time: real] NORMALTTY [origin: string, destination: str stop: string, time: real] LETTER PRlORrrY [origin: string, destination: stri length: {real}, width: {nsA}, level: integer, time: rral, area{real}] BOX NORMAU TV [origin: string, destination: strii length: {real}, width: {real}, height: {real}, stc^: string, timi real, volume: {real}] X dupe cho Bing (C, =>), me vi p dugc cho nhu Hinh Chiing toi luu y ring, Bing 1, cic phucmg phap ciia cic ldp cd thi dugc djnh nghTa mpt cich tudng minh dya tren mpt chiln lupc hpi xic suit ® ([7], [9]) d l xac djnh vi tinh toin xic suit ciia cic gia trj cd the tri ve ciia chiing Ching ban phucmg phap area cac ldp LETTER vi LETTER_PRI0RITY cd thi dupc djnh nghTa nhu sau: area((F,, at, fit): {real}, {Vj, aj, fii): {real}): ((F, a ^ : {real}) 110 T(C) yjv^u HQi thto ICT.rda'06 Proceedings of ICT.rda'06 Hanoi May 20-21.2006 Tuy nhien, luge dugc djnh nghTa nhu t i ^ cd thi li khdng nhdt qudn (consistent), nghia li khdng phii ludn ludn tim dugc mpt tip cic doi tugng thda mpt phep gin xic suit vi phan loai cic Idp dugc bilu dien bdi dl thj {C, =>) vi phin hoach cic canh Mpt lupc FPOB li nhit quan nlu va chi neu nd c6 mpt mo hinh nhu djnh nghTa sau day Dinh nghia 2.5 Gii siir S = (C, x, =>, me, p) li mpt luge FPOB Mpt dien dich cua S li mpt anh xa tir C din tap cic tip eon hiru h^n ciia mdt tip O Mot dien djch e cua S la mpt mo hinh ciia S nlu va chi nlu: e{c)*0,yceC e(c) n e{d) = 0,\/ c,d e C vai c wad phan bift va c, d e P e me{C) THITA KE VA Djnh nghia 3.1 Gia sir S = (C, x, =>, me, p ) la mdt lupe FPOB vi O la mpt tip danh hifu doi tupng, mdt thi hifn FPOB tren S li mpt cap {n, v) do: ;r: C -> li anh xa dit tuong iirng moi ldp c thupc C vdi mpt tip hiru ban ciia tap O, cho ;i(ci) n ;z(c2) = 0, V Ci 5t C2 e e(c)c e{d), \I c,d e Cvaic=>d \e{c)\= p{c,d).\z{d)\'^c,deC c=>d mpt CO sd ddi tupng phii hgp vdi cic kilu dir lifu va chiln luge thira kl tren luge niy Mpt thi hifn tren lupc FPOB cho thay dugc tinh trang, quan hf thiira kl ciia mpt tip doi tupng FPOB Thi hifn ciia eo sd doi tugng xac suit md dugc djnh nghTa nhu sau: vdi CAC THfe HI?N FPOB Doi vdi FPOB, dl giai quylt moi quan hf thira kl ndi chung vi da thira kl ndi rieng ciia cic ldp va Idp cha, chiing tdi ap dyng ciing mpt chiln luge thira kl nhu [7] Cho mpt lupc S = (C, x, =>, me, p ) ip dyng chien luge thira ke niy tren S se dan den mpt lupe mdi S* = (C, x*, =>, me, p) chi khic S d phep gin kieu x* Cy the vdi mdi c G C, x*(c) = [At x{dt)At Ak x{dt,).Ak], dd A , Ak cic tinh chit mire cao nhit dugc thira ke bdi c thong qua chiln luge thira kl tir cic ldp cha d\, , dk tuong irng Chiing ta noi S* la sy thira kl hoan toin ciia S Mpt lupc dd S dugc gpi li thira kl diy du neu va chi neu S = S* Ke tir bay gid chiing toi gii siir rang tat ca luge FPOB la nhit quan va dugc thira ke day dii Cho mpt luge S, mpt the hi^n (instance) FPOB tren S dupc dinh nghTa nhu C Ngoai anh xa ;i* : C -^ 2*^ dupe djnh nghTa bdi ;i*(c) = u{;!(c')| c'eC, c'=^* c } la tap cac doi tupng thupc ldp c (cae doi tupng c hoac ldp con, chau eiia c) V la anh xa dit tuong irng moi o e ;z(C) vdi mpt gii trj bp xic suit md kieu x(c) cho o G 7i(c) Vi du 3.1 Mpt thi hifn I = {n, v) tren lupc FPOB vi dy 2.4 dupc chi Bing vi Bing Trong dd, about ={1: 0.3, 2: 1, 3: 0.3}, about_6 = 3xabout_2 = {3: 0.3, 6: 1, 9: 0.3} vi about_6.2 = 3.lxabout_2 = {3.1: 0.3, 6.2: 1, 9.3: 0.3} la cac so md thi hifn dien tich khong chinh xic cua mpt li thu, dupc tinh theo phuomg phap area vi dy 2.4, dya tren tich hai so mo theo cdng thiire sau: la {A B){z) = sup^ ^.ymin[A{x), B{y)], vdi mpi so thyc x, y vi z Cae khoang xic suit cho gii trj tri vl cua phuong phap aree dugc tinh theo chien luge hpi dye lap ®,„ ([7] [9]) Trong eo sd doi tupng truyen thong pham vi ciia mpt Idp bao gom tat ea cae tugng thupc ve Idp Trong FPOB, ph?in v xac suit cua mpt ldp chi xac suat dc che mdi d6i tuong thupc ve lap Noi each khac Proceedings of lCT.rda'06 Hanoi May 20-: Ky y^ HQi thto ICT.rda'06 ph^m vi xic suit ciia mpt ldp li ham xic djnh cac thinh vien khdng chic chin ciia ldp nhu djnh nghTa 3.2 n(c) Neu o G ;i*(c), thi ext{c){o) = {1 Neu o G 7r*{d) vke{d)r^e (c) = mpi md hinh E cua S thi ext{c){o) Bang 2: Anhxan vin* c tuomg irng mdi o e niQ vdi mpt tap hu'u ti khoing [0, 1] cho: 7I*(C) PACKAGE {0,} { | , O2, O3, O4 } LETTER {} {02,03} BOX {} {04} TUBE {} {} PRIORITY {} {02,03} NORMALITY {} {04} LETTER_PRIORITY {02, 03} { , } BOX_NORMALITY {04} {04} Ngugc \a\,ext{c){o) = {p\ plati cac xic suit cic canh tren mdt du c din mpt ldp rf G C, a dia l< nhit cho oe 7i*{d) va c =>* d] Vi du 3.2 Gii six I la thi hifn cii dd FPOB vi dy 3.1, pham vi xi cua cic ldp LETTER_PRIORITY B0X_N0RMALITY ddi vdi 0| va 02 nhu s eA:/(LETTER_PRIORITY)(o,)= {C exr(BOX_NORMALITY)(Oi) = {0 eA:/(LETTER_PRIORITY) (02) = { Bang 3: Phep gan v oid Ol 02 03 04 v(oid) [origin: ({Hanoi}, u, u), destination: ({Saigon}, u, u), time: ({26, , 30}, u, u>] [origin: ({Saigon}, u, u), destination: ({Hue}, u, u), length: ({3,3.1}, O.Su, I.Su), width: ({ about 2}, u, u), level: ({l},u,u),time:({4,5},0.8u,I.2u), area: ({about 6, about 6.2}, O.Su, I.Su)] [origin: ({Hanoi}, u, u), destinatkin: ({Nhatrang}, u, u), length: ({2,2.1}, u, u), width: ({1, l.I},0.6u, I.Su), level: ({2},u, u), time: ({4, 5, 6}, 0.6u, I.Su), area: ({2,2.I,2.2,2.3I},0.6u, I.Su)] [origin: ({Saigon}, u, u), destination: ({Hanoi}, u, u), length: ({3,3.1}, 0.6u, I.4u), width: ({2}.u, u) Height: ({4}, u, u), stop: ({Nhafrang, Hue}, O.Su, I.6u), time: ({SO, SI, 82}, 0.6u, I.Su), volume: ({24,24.8}, 0.6u, I.4u)] Djnh nghia 3.2 Gii sir I = (;r,v) la mpt thi hifn tren mpt lupc dd FPOB S = (C, x, =>, me, p) Vdi mdi Idp c G C, pham vi xac suit ciia c, dugc ky hifu ext{c), la mpt inh X9 dit eTr(BOX_NORMALITY)(02) = {0 TRUY VAN THONG QUA PHUa> PHAP Vugt qua cac h^n che cua md hinh [7], md hinh dugc de nghj khdng chi cho bilu dien hanh vi vi cac thao tic tren ca tugng (nhu cic md hinh lip trinh h doi tugng) bdi cac phucmg phap, mi cdi phep truy vin ddi tugng thdng qua sy thi cua chiing Phin niy trinh biy mdt sd ap truy vin xac suit md ddi tupng qua phep (selection operation) bing each ap dung phuomg phip Trudc hit, cic khai niem diin dich sudt (probabilistic interpretation) ciia thiire chpn, dieu kien chon ma {fuzzy selec condition), dien djch xic suit ciia dieu chpn vi phep chpn dudi diy dya tren nghTa tuong img [7], vdi md rpng la bilu thixc dudng di bilu thuc chpi diiu kifn chpn cd sy tham gia ciia cic phu phap Djnh nghia 4.1 Gia su S = (C, x, =>, p ) la mpt lupc FPOB, I = (;r, v) la the hifn tren S, x la mpt bien ddi tugng v i n{C) Dien djch xic suit theo S, I va o, d bilu thj bdi probs.i. aboutjS) dugc tinh toin tir djnh nghia 4.1 probs,i^2(Jf area - * aboutjS) = aboutjS) = [0.36, 0.81] c [0.3, djnh nghia 4.3 probs.1^2 1= l thda r tra tuong ty, thay ring khdng cdn nao khac thda dieu kien chpn [0.8x u{about_6) x pmh{about_6 -> Vi du 4.5 Cung the hifn I tn aboutjS) + 3.1, truy vin "Tim tit ca cac hop due 0.8 X u{about_6.2) x pTah{about_6.2 ->thdng thudng va cd the tich Idn he aboutjS), xac suit 30-80%" cd thi dugc thyc min{l, 1.8 x u{aboutjS) x prob(a6oMr_ aboutjS) + r = a*(I) 1.8 X u{about_6.2) x prcA){about_6.2 -> vdi | = (x G BOX_NORMAl aboutjS))] A:.volume(jc.length, x.width, jr.hei = [0.8 X 1/2 X 0.90 + 0.8 x 1/2 x 0.0, 24)[0.3, 0.8] Chi cd ddi tugng 04 tl m/;i(I, 1.8 X 1/2 X 0.90 + 1.8 x 1/2 x 0.0)] = truy vin niy [0.36,ffH>i(1,0.81)]= [0.36, 0.81] That viy, volume(o4.length, 04 Vi probs,i.„2(jcarea^ aboutjS) = [0.36, 0.81] c [0.3, 1.0] nen 02 thda diiu kifn chpn da neu 04.height) = ({24, 24.8}, 0.6M, probs,i^(jr.VOlume > 24) = [0.3, 0.7 dang thiy probs.i.o4(JC e BOX_NORMA [1.0, 1.0] Do dd probs.i.o4' BOX_NORMALITY 24) 0.7] c [0.3, 0.8] Con cac doi tugn khdng thoa diiu kifn chpn Chung tdi luu y d diy ring prohiaboutS -> aboutjS) = 0.90 v i prob(adoM/_d.2 -> about_6) = dugc tinh toin theo [7] nhu di noi d tren D|nh nghia 4.4 Gia sir S = (C, x, =>, me, p) la mpt luge dd FPOB, I = (;i; v) la mpt thi hifn tren S vi i|) li mpt diiu kifn chpn md tren biln doi tupng x Phep chpn tren I theo (|), dupc ky hifu a»(I), li mpt thi hifn I' = {n' vOtren S, dd: 7t'{c)={oe KET LUAN Trong bii bio nay, chiing tdi d thifu mpt md hinh co sd ddi tugng xi md la md hinh [7] bing each tic cac phucmg phip vio cac ldp ddi t Chiing tdi da thdng nhit khii nifm thud v i phuong phap khai niem chung I chat, sau kieu bp md dugc md rpng dt nghTa mpt ldp cic doi tugng thdng qu tinh chit cua chung Tilp theo, cac khai [7] dugc md rpng mdt cich tuong lim CO sd dl thyc hifn cac truy van c dyng cic phuong phip n(c)| probs,i.o 1= • } vi v'= VI n'{Q (nghia li inh x? v thu hep tren n'{Q) Vi dy 4.4 Gia sii I = (;z; v) li thi hifn ciia luge S vi dy 3.1, thi truy vin "Tim tat c i cic gdi cd difn tich li aboutjS vdi xic suit it nhit li 30% va dugc vin chuyin khdng qui gid vdi xic suit it nhat li 80%" cd thi dugc thyc hifn bdi phep chpn Sy tich hgp phuomg phap vao cic \a\ tugng cho phep md hinh hoi cic ap dyn; hon Ngoai ra, cac truy vin doi tugng I cich thyc thi cac phuong phap cdn cho \ dat dugc cac thdng tin chua cd sin o dvt lifu trudc dip ung cic ddi hdi vir thdng qua gii trj cua cac doi tugng r = a4(I) vdi (|) = (x.area(jt.length, jc.width) -^ aboutj6)[0.3, 1.0] A (xtime ^ 5)[0.8, 1.0] Thi hifn I' tren S gom chi mpt doi tupng 02 Thit viy, vdi 02, tur vi dy 4.1 va 4.3 thi probs.i^2(x.time < 5) = [0.8 1.0] c [0.8, 1.0] vi probs,i^2(jfarea(x.length, Trong cic bude tilp theo, chung to xay dyng cac phep toin dai so doi tugng i tich Cartesian, chilu, kit, hgp, giao vi tru x.width)-> m H6i thto ICT.rda06 ia tham gia cua cac phuong phip Diy sg la pa sd cho mpt ngdn ngii truv vin doi tupng 1,0^ chinh tren mpt co sd doi tupng xic suit nid thyc tl Tii lieu tham khao Proceedings of ICT.rda'06 Hanoi May 20-21, 2006 Soft Confuting, Physica-Vcriag, 1999, 26, 41 64 [12] Van Gyseghem N., De Caluwe R The UFC database model: dealing with impeifec, information In De Caluwe, R Fuzzy anc Uncertain Object-Oriented Databases: Concept and Models, World Scientific, 1997,123-185 [I] Baldwin J.F., Martin T.P., Pilsworth B.W Fril Ve cac t i e gia: Jiezy and evidential reasoning in artificial intelligence Research Studies Press, 1995 [2] Berzal F., Marin N., Pons O., Vila M.A A framework to build fiizzy object-oriented capabilities over an existing database system In Ma, Z (Ed.): Advances in Fu2zy ObjectOriented DataJjase: Modeling and Applications Idea Group Publishing, 2005,177-205 [3] Biazzo V., Giugno R., Lukasiewicz T., Subrahmanian V.S Temporal probabilistic object bases IEEE Transactions on Knowledge and Engineering, 2002,15,921-939 Thac sT Nguyen Hoa tot nghiep Dai hoc Sir pham Vinh nganh Toan nam 1982, nhin bing Thac sT Khoa hpc Miy tinh tai Dai hoc Bach Khoa Thanh H6 Chi Minh nam 2003 Hien dang lam nghien ciru sinh tai Dai hpc Bach Khoa Thanh Ho Chi Minh Th.S Hoa dang la giang viln [4] Bordogna G., Lucarella D., Pasi G A fiazy object-oriented data model Proceedings of the IEEE Intemational Conference on Fuzzy Systems, 1994,1,313-318 tnrdng Dai hpc Ma Thanh ph6 H6 Chi Minh [5] Cao T.H., Rossiter J.M A deductive Nhihig ITnh vyc quan tam nghien curu cua Th.S probabilistic cmd jiizzy object-oriented database Hoa gom c^c Co sd DQ lifu Xic suat-Md va Tinh language Fuzzy Sets and Systems Elsevier toan Mlm Science,2003,140,129-150 E-mail: hoa-hanh@hcm.vnn.vn [6] Cao T.H., Nguyen H Towards Jiizzy and probabilistic object bases Proceedings of the 3Td Tien sT Cao Hoing Try International Conference on Intelligent tot nghifp D^i hpc Bich Technologies and the 3ni Vietnam - J^KUI Khoa TPHCM nganh Symposium on Fuzzy Systems and Applications, Khoa hpc & YJj thuit Hanoi-Vietnam, 2002,35-41 May tinh x&m 1990, nhin [7] Cao T.H., Nguyen H Fuzzy and probabilistic bing Th^c sT va Tiln si object bases In Ma, Z (Ed): Advances in Fuzzy nganh Khoa hpc M^y tinh Object-Oriented Databases: Modeling and lan lupt tai Vifn C6ng Applications Idea Group Publishing, 2005, 46nghf A chiu nSm 1995 va S4 Dai hoc Queensland nam [8] Cross V Toward a unifying framework for a 1999, lam nghien ciru sau fuzzy object model In Proceedings of the 5*' Tien sT tai IEEE Intemational Conference on Fuzzy Systems, 1996, 85-92 D^ii hoc Bristol va Dai hoc California-Berkeley tii [9] Eiter T., Lu JJ., Lukasiewicz T., Subrahmanian nam 1999 din 2001 Ts Tru hifn la Pho giio su, V.S Probabilistic object bases, ACM Pho trudng Khoa Cong nghf Thcng tin, tnrdng Transactions on database systems, 2001, 26 Dai hoc Bach Khoa TPHCM, vien Hpi ding 264-312 Bien tap Tgp chi Phdt trien Khoa hoc vd Cdnfi [10] Nguyen H., Cao T.H Fuzzy-probabilistic object nghe DHQG-HCM Cac llnh vyc nghien ciru gom bases and the selection operation Journal of Tinh toan Mlm va Web co Ngff nghTa Computer Science and Qrbcmetics, 2004, 4, Email: tru@cse.hcmut.edu.vn 329-342 [II] Yazici A., George R Conceptual modeling of complex and uncertain information-Fuzzy database modeling Studies in Fuzziness and