Deductive model for active database systems

Researches on deductive database systems are not new ones, but propose of data model allowing to manipulate data and knowledge on a particular framework is not [r]

VNU JOURNAL OF SCIENCE, Mathematics - Physics, T.XXIl, N04 , 2 0 06

DEDUCTIVE MODEL FOR ACTIVE DATABASE SYSTEMS

Do T r u n g T u a n

College o f Science, V ietn a m N a tio n a l University, H anoi

N g u y e n T h i T h a n h D u y e n

G raduated S tu d en t, V ietn a m N a tio n a l University, H anoi

A b s tra c t Researches on deductive database systems are not new ones, but

propose of data model allowing to manipulate data and knowledge on a

particular framework is not easy but interesting goal The paper aims at a

model for knowledge in educational and training environment, beside of a

presenting the achievement in the existing systems concerning knowledge

Certain data manipulation techniques for knowledge acquisition are proposed

in the model, as knowledge discovering techniques in deductive database

systems An active database is proper for applying such technique on rule-based

knowledge

Index Terms—Rule, active database, deductive, datamining

1 I n t r o d u c t i o n

A rtificial intelligence h a s know ledge as a n im p o rta n c e object U n d e r s ta n d in g and u sin g knowledge is ob tain ed goal of h u m a n society a n d scientific research Since 1964, ta c it knowledge and o th e r k in d of know ledge is e x am in ed in th e

relatio n sh ip to ta c it knowledge In th e r e s e a r c h on d a ta b a s e system , know ledge is rep re se n te d in sy stem s Among proposed d iffe re n t k in d s of knowledge, ru les are preferred

B eg in n in g w ith file m a n a g e m e n t sy s te m s in 1960, a lot of app licatio n sy stem s based on h ie ra rc h ic a l model, netw ork m odel (1960, 1968), re la tio n a l model (1970) have re m a rk a b le role in th e economy society A fter 1990, ad v an ced model, such as distrib u ted model, object o rien ted m odel a n d d e d u ctiv e model h a s studied The

s y s t e m s b a se d on th ese advanced m odels did n o t d is tin g u is h e d from ones of the second g e n e ra tio n of d a ta b a s e m a n a g e m e n t sy ste m s

In th e years 80’s, some works approached knowledge m anipulation ir nformation system s A solution for lose gap or stric t gap betw een artificial intelligence and database th eo ry has some results for knowledge m an ip u latio n [5, 11]

R e se arch e s on logic p ro g ram m in g allow s to prove th e o re m s au to m atica lly , t(

set a rela tio n sh ip b etw een ev en ts and d ed u ction by (i) p roving theory; and (ii) mode

'heory The p roving theory p re se n ts sp ecificatio n s of t r u e re a so n in g a fte r p rem ise s the model th e o ry p re s e n ts a sse rtio n s a f te r e v en ts Prolog la n g u a g e u se s Horr clauses a n d b a c k w a rd c h ain in g inference w ith s e m a n tic s of both th e proving theor;

42

D edu ctive M odel fo r A ctive D a ta b a s e S ystem s

43

and th e model theory T his la n g u a g e su p p lie s capacities of (i) e v e n t r e p r e s e n ta t io n ' (11) inference; (ill) rec u rsiv e re a so n in g ; (iv) u p d a tin g in te g rity c o n s t r a i n ts T he system XBS [7] is a success w ith th e HiLog, a developm ent on Prolog

At th e point of d a ta b a s e re s e a rc h , t h e r e is a sim ila r of re la tio n a l d a ta b a s e system s a n d logic p ro g ram m in g T r a n s f e r r in g a p red icate cla u se to a r e la tio n a l clause is not difficult H ow ever it e x ists problem s, such as (i) effectiveness of

organ izin g a g rea t am ou n t of k n ow ledge; (ii) rea so n in g cap acity of query la n g u a g e

in the re la tio n a l model T he re la tio n a l q u e ry lan g u a g e can not exhibit whole application; so m etim e it n e ed s a p p lica tio n p ro g ram s Some ded u ctiv e m odels focus

on Prolog a n d tec h n iq u e for r e la tio n a l d a ta m a n ip u la tio n D atalo g is a r e s tr ic t Prolog, u s in g n e g atio n a n d is a a p p ro a c h in d eductive model of y e a r 90’s

C oncerning to th e a sp e c t of e n d -u s e r in te rfa c e in deductive d a ta b a s e s , it ex ists problem s [5] (l) ded u ctiv e d a ta b a s e s use flat s t r u c tu r e t h a t is not a p p ro p r ia te to exhibit complex objects; (ii) c ap acity of m od elin g objects; (iii) d a ta schem a a t the user level; (lv) u p d a tin g know ledge In a n a ly z in g th ese problem s, th e r e a re solutions as follows :

1 D eductive la n g u a g e for complex objects D atalog is developed in o rd er to specify complex d a ta s t r u c tu r e , n e s te d d a ta It o b tain s (i) LDL; (ii) COL-(iii) HiLog; a n d (iv) R elationalog;

2 Object o rie n te d d ed u ctiv e lan g u a g e T his a p p ro ach ex am in es (i) objects-(li) complex d a ta ; (iii) m eth o d s; (iv) class; (v) h e ritag e; a n d (vi)

e n c a p s u la te Some r e s u lts a r e (i) O-logic; (ii) F-logic; (iii) ROL; a n d (ivi

3 D a ta sch em a a t u s e r level T h e second g e n e ra tio n of d a ta b ase

m a n a g e m e n t sy s te m s w ith r e la tio n a l d a ta model h a s not been clear on sch em a specification w h e n u s e rs w a n t to in tro d u ce d a ta s tru c tu re

in te g rity c o n s tr a in ts T h e f ir s t o rd e r p re d ic a te logic a n d h ig h e r tools may

be used for solving th e problem T h e la n g u a g e allow ing to specify user sch em a (i) HiLog; (ii) L2; (iii) F-logic; a n d (iv) ROL;

4 U p d a tin g knowledge D a ta b a s e a p p lic a tio n s need a n in te ra c tiv e interface for u p d a tin g m e ta d a ta U p d a te objects are (i) extensive d a ta b a s e s and

in te n tio n d a ta b a s e s ; (ii) u n d e fin e d d a ta ; (iii) collection, set; and (iv)

u p d a te a fte r conditions

2 A r c h it e c t u r e o f d e d u c t i v e m o d e l w i t h a c t iv e d a t a b a s e s

R e l a t e d w ork

Some system s focus on integrity constraints in database management

system s for m a n ip u la te k now ledge in d a ta b a s e M eta d a ta in th e d a ta dictionary composed of design ru les a r e a k in d of know ledge in d a ta b a s e T h e system s in [6 7] are p ro to ty p e for m a n ip u la tin g know ledge on in d a ta b a s e Besides m eta data the

m0ng d a t a let th e sy stem s to discover association ru le [1, 2, 4]

relationship “ ^ ^ a t a re la tio n sh ip a re a cc u m u late d as knowledge

Inform ation ta lja ses 9) 10j u s e m e c h a n is m s o f (i) ev en t; (ii) condition; a n d (iii)

Active ^ ^ ace in te g rity c o n stra in ts, policies u n d e r th e form of knowledge, actions to in n ge a detector a n d a trig g e r for realizing actions w h en th e

TlieJ conditions are 5t!ãré satisfied

/ fitr d e d u c tiv e m o d el w ith an a c tiv e d a ta b a s e

otiet for deductive d a ta b a s e s h a s n o t yet s e p a ra te d from relational

D a t 8 d a r t i f i c i a l system s T h ro u g h o u t, i t i s necessary t o construct an

rr.odil w , -u ctive model

independent decuuu w

F ternal level allows to specify sy stem re q u ire m e n ts , to q uery to 1* T^uctxve d a ta b a s e (goal) The process of in p u t d a ta a n d knowledge is

Used in th is level In fact, i t needs in te rac tiv e interface for gffrring th e u se r d e m a n d s to a p p ro p ria te query to the (data/ j^velfdge) m a n a g e m e n t system s;

r n i j e n e n t level composes of d a ta m a n a g e m e n t a n d knowledge

2 na^ement A app ro ach u sin g Prolog inference engine is good; w h a te v e r ceis a c o rre sp o n d e n t d ev elopm ent to scale of deductive model, with

e 's tp p a n d h ig h e r level specification ;

y^ysi«a level is c o rre sp o n d e n t to o rg an iz atio n of events a n d rules Rule

' Ted know ledge is proper w ith in in feren ce en g in e in m an agem en t level

' fne )ther kinds of knowledge are transferred to rules.

Tuan DoTrung, Nguyen Thi T hcnhD uyen

44

ấ

user

external level

1 1 knowledge ; N

m anagem ent !

« I l J

1

• 1 physical level i o

Q

Inference engine Prolog

knowledge b a se

Figure 1 D eductive model w ith 3 levels

JỈ diUsi* the a r c h ite c tu r e

, ,.n s are som e modules in th e a rc h ite c tu re of deductive model in active

r Q lO V 11!

d a f b *

â,áen‘-D e d u c tiv e M odel fo r A c tive â,áen‘-D a ta b a se Systems

3

4

5

45

1

2

Query m odule It allows to describe events, d a ta an d know] rỉ end u s e rs in terface, d a ta stru c tu re s and knowledge, integrity Ir‘ the

D a ta m a n a g e m e n t module This module supplies tools f<0> se

manipulating data via query language Query concerning k nể and

tr a n s f e r r e d to p red icates in Prolog D a ta m a n a g e m e n t I1( is

dictionary, softw are package for d esign purposes; a Cat£

Q uery la n g u a g e module The existence of th is m odule is f0l ,

specification U se r d e m a n d s for reaso n in g on event, k n o w ltd g Ị

exist in re la tio n a l language Therefore, it is ex am in e d by ^

th e n t r a n s f e r r e d to a Prolog goal M a tch in g re la tio n a l algeb s ao d u le>

M etab ase D a ta b a s e is tra d itio n a l one Knowledge b a sỊ jk

physically as d a ta b a s e W h atev e r a un iq u e m e ta b a se for ho h 1 0ỉ^*niỉeđ

At th e level of m a n a g e m e n t, d a ta and knowledge a re d is tin ru ilh eierr,d* Knowledge discovering module In th is m odule, d a ta m in iip ]

association rule discovering m ethods a re proposed W ih , Ị ì f rĩtỉm s’

event-condition-action,

knowledge base

new re la tio n sh ip am ong d a ta Ip of

*ditd t0

ị

q u e r y , g o a l

lin g u is t ic s

q u e r y l a n g u a g e

n

I

B ị

d a t a d ic t io n a r y •

d a t a m a n a g e m e n t (r e la t io n a l m o d a l)

- $ ■

in te g rity c o n s t r a in t s

s p e c ific a t io n

d a l a m in g m e lK o d

a s s o c i a t io n ru le

k n o w le d g e m a n a g e m e n t ( P r o lo g » 0 9 10 « )

R u le {R i}

Figure 2 A rc h ite ctu re of th e deductive model w ith active dataj'g

B esides m odules, it is n e ce ssa ry to im p lem en t form al specificaio-1

no tatio n , sy n ta x a n d se m a n tic asp ects specification A box for ev>rt • ' n^ ete action m ay be e ith e r (i) a n e x te rn a l application; or (ii) a function eial'0 ri tl5r‘‘

3 C o n c lu s io n s

T h e p a p e r p r e s e n ts th e principal modules of th e arch ite ctu re o' (

model in active d a ta b a s e s T h e d a ta b a s e m a n a g e m e n t system is relat}ra u ^ e knowledge m a n ip u la tio n b ased on Prolog inference engine

46 T uan D oT rung, Nguyen Thi Than hD u yen

K now ledge is in in te g rity c o n stra in ts , a sso c ia tio n ru le s am ong d a ta , events a t

e x te r n a l level In o rd er to a cq u isitio n know ledge, d a ta m in in g a lg o rith m s, asso ciatio n ru le discovering are applied

R e f e r e n c e s

1 B o tzer D 0 Etzion, S e lf-tu n in g of r e la tio n sh ip s am o n g ru le ’s

c o m p o n e n ts in active d a ta b a s e s sy s te m s, I E E E tran On knowledge a n d

d a ta en g in e erin g , V 16, N.3(2004), p 375- 379

2 F lesca s s Greco, Declarative se m a n tic s for active rules, 2003, 27 p.

3 L aw Y H W ang, c Zaniolo, Q u e ry la n g u a g e s a n d d a ta models for

d a ta b a se seq u e n c es and data str e a m s, Proc o f the 30th V L D B conference,

2004, 12 p

4 Lee c M Chen, Progressive p a r t i ti o n m in e r : a n efficient a lg o rith m for

m in in g g e n e ra l tem p o ral asso ciatio n ru le s , I E E E tran On knowledge a n d

d a ta en g in e erin g , V 15, N 4(2003), 1004- 1017

5 Liu M D eductive d a ta b a s e la n g u a g e s : pro b lem s a n d solutions, A C M

c o m p u tin g surveys, V 31, N 1(1999), p 27-59.

6 Liu M Design an d im p le m e n ta tio n of th e ROL sy stem , J o u rn a l o f

in te llig en t in fo rm a tio n system s, N 14(2000), p 1-21.

7 S a g o n a s K., T Swift, D s W a r r e n , XSB as an efficient deductive

d a ta b a s e engine, Proc O f S I G M O D , 1994, p 442- 453.

8 S u d e r m ie r A., D ietrich S u z a n n n e w , S h a h V., A n active database

app ro a ch to integrating black-box so ftw are com ponents, A rizona sta te

U n iv e rs ity , 1998

9 U r b a n s D e t al., An e v a lu a tio n of d i s t r ib u te d co m p u tin g options for a

ru le -b a s e d a p p ro a ch to black-box so ftw a re co m p o n en t in te g ra tio n , Proc o f

3 rd int workshop a d va n ced issues o f e-commerce a n d web-based,

in fo r m a tio n system s (W ECW IS’01), 2001.

10 U r b a n s D e t al., D elta A b stra ctio n s: a te c h n iq u e for m a n a g in g d a ta b a s e

s t a t e s in r u n tim e debugging of activ e d a ta b a s e ru les, I E E E tran On

k now ledge a n d d a ta en g in eerin g , V 15, N 3(2003), p 597- 612

11 Y an g B et a l ( D eductive sy n th e s is of workflows for e-science, IE E E inter

S y m p o s iu m on cluster c o m p u tin g a n d the grid., 2005, p 168- 175.

12 Zaniolo c H W ang, Logic-based u se r-d e fin e d a g g re g a te s for th e next

g e n e ra tio n of d a ta b a s e sy ste m s, in “T h e logic p ro g r a m m in g p ara d ig m : a

25 y e a r p ersp ectiv e”, Ed S p rin g e r, p 401-426

13 Zaniolo c , M in in g databases a n d d a ta stre a m s w ith query languages and

r u le s , 14 p., 2005.