S. Manchanda and D. S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. Morgan Kaufmann Publishers, San Mateo, CA, 1988.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....section, we compare DatalogU with several related proposals. Some critical analysis of these proposals can be found in the survey [17] One of the novel features of DatalogU is its natural support for set oriented or bulk updates. First, let us see how other proposals handle bulk updates. In DLP [22], the bulk updates have to be represented by using recursive update rules. For example, the following rules in DLP can be used to give every employee a R salary increase: hraise(R)i : emp(E;D;S) hraise(R)i : hemp(E;D;S)i(S = S (1 R)hraise(R)i(h emp(E; D;S)i) Updates in DLP have an ....

....the implementations of ROL [14, 15] ROL2 [18, 19] and Relationlog [16, 28] There are several important issues which we would like to address. Being a database oriented language, DatalogU does not directly support non determinism updates that are important in logic programming based approaches [22, 5, 9, 30]. In LDL [26] the choice operator is used to provide non determinism in deductive databases. We would like to investigate if the explicit use of the existential quanti cation provides a natural way to support non determinism in deductive databases. Also, we intend to extend DatalogU to support ....

S. Manchanda and D. S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundation of Deductive Databases and Logic Programming, pages 363-394, Los Altos, CA, 1988. Morgan Kaufmann.

....can be performed at definition time to execute update requests more efficiently. 1 Introduction The theory of database updates is currently receiving much more attention than in the past [1] and recently, the problem has been largely studied by various authors with regard to deductive databases [7, 12, 19, 23, 25, 29, 32] and related frameworks [13, 35] This study is particularly important since the main goal of deductive database research is to integrate logic programming and databases in order to provide future database systems with more powerful mechanisms for data representation and computation [8] and sofar ....

....use only the rules and we show that a computation rule suffices and can be freely chosen. Finally, the concept of completeness of updates is defined with respect to operational issues, we give instead completeness results with respect to the declarative semantics. Several other interesting papers [7, 16, 17, 23, 25] deal with issues related to the problem addressed in this paper, but they present some differences that do not allow a quantitative comparison. In [7] an intensional update is considered as a general 36 property, defined by means of a formula involving quantifiers and meta predicates, specifying ....

[Article contains additional citation context not shown here]

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394, Morgan Kauffman, Los Altos, 1988.

....end consistent. If the set is not consistent the trensection is aborted end no updete in the set is performed. The notion of consistency is en iraportent one, in thet it prevents e set of updetes conreining both en insertion end e deletion of the seine feet to be executed. By contrest in DLP [47], D [54] Tq [20] end DL [5] updetes ere executed es soon es they ere evelueted. This epproech leeds to complex sementics end to computetions performed in e sequence of stetes 12 instead of in a single one. In the following we recall the language and its informal behavior (see [53] for a ....

S. Manchanda and D. S. Warren. A Logic-based Language for Database Updates. In J. Minker, editor, Deductive Databases and Logic Programming, pages 363 394. Morgan-Kaufmann, 1987.

....attributes, because a derivation rule expressing an attribute value in terms of a number of others is just a special kind of integrity constraint. The communication defining the simultaneous update to the related attributes is then a kind of update procedure as defined by Manchanda and Warren [37]. Further work needs to be done on this, both at the specification and at the operational level. For example, some global constraints require communications that are broadcast to a set of one or more objects that is not yet known at specification time. For example, in ComplexPersonObjects, C5 ....

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363394. Morgan Kaufmann, 1988.

....under arbitrary sequences of update transactions. A wide variety of proposals for this exist in the literature (e.g. Abiteboul [1] Grahne [13] Katsuno and Mendelzon [20] Winslett [48] Fagin, Ullman and Vardi [10] Ginsberg and Smith [12] Guessoum and Lloyd [16, 17] Manchanda and Warren [32], Kowalski [22] Bonner and Kifer [6] In this paper, we advance a substantially different approach. To begin, we take seriously the fact that, during the course of its evolution, a database will pass through different states; accordingly, we endow updatable database relations with an explicit ....

....is true. This makes historical queries possible (Section 7.1) and provides for a theory of integrity constraints (Section 6. 1) This is impossible or extremely awkward to do within the logic for those approaches to updates formalized in modal logics e.g. dynamic logic, Manchanda and Warren [32], temporal logic, Casanova and Furtado [8] or in path based logics (e.g. Bonner and Kifer) for which there is only an implicit notion of state. Logical Status of Transactions A feature of the situation calculus, related to that of states as first class citizens, is that transactions and ....

[Article contains additional citation context not shown here]

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. Morgan Kaufmann Publishers, Inc., Los Altos, CA, 1988.

.... at least two di#erent topics: An update and a transaction language should be provided and typical atomicity, consistency, isolation, and durability (ACID) properties have to be guaranteed [24] This requirement has led to the definition of several languages integrating logic and updates [1,2,9,10,15,16,31,32,35]. In general, all those proposals are based on including special atoms denoting updates in (typically Datalog) rules. The proposed approaches di#er for several aspects, such as the semantics assigned to the resulting language (declarative vs. operational) the evaluation techniques, the update ....

S. Manchanda and D.S. Warren. A Logic-Based Language for Database Updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. Morgan-Kaufmann, 1988. Integrating Active and Deductive Rules 133

.... # 0 2 ;S 0 2 p(X) #2 = fX=ag S2 = hf Gammaq(X)g; i # 0 2 = fX=bg S 0 2 = h; f Gammaq(X)gi Fig. 3. Obj inh Datalog proof for transaction T2 of Example 7 Inheritance in a Deductive Object Database Language with Updates 19 bodies) are executed as soon as they are evaluated [MW87] Under this assumption the evaluation of a rule is performed in a sequence of states and thus the declarativeness of the query part is lost. Under the marking and update phases, the first phase is declarative and preserve this property for the query part while accommodating update specification ....

S. Manchanda and D. S. Warren. A Logic-based Language for Database Updates. In J. Minker, editor, Deductive Databases and Logic Programming, pages 363--394. Morgan-Kaufmann, 1987.

....rolled back, but only those that have erroneously been executed since the last choice point. Due to space limitations we cannot go into detail here, but refer the interested reader to [FWF00] 5 Related Work 5. 1 Logical Foundations of Updates The language DLP proposed by Manchanda and Warren [MW88b] is based on dynamic logic [KT90] although the programs within the scope of a modality operator are atomic. Instead, complex operations are specified by update rules for which a model theoretic semantics is presented. Like in U Datalog [MBM97] the basic operations are restricted to insertions ....

....work. It forms a modal logic [Eme90] for the representation of changes in which the sequential composition of operations is handled explicitly and arbitrary basic operations 41 can be integrated. The interpretation of formulas is defined w.r.t. state paths. This generalizes the semantics of DLP [MW88b] that is based on pairs of states. The general logical concepts have been restricted to form a rule based language. This language has a model theoretic and a proof theoretic semantics, where multiple answers for the same update query are interpreted as non deterministic solutions with specific ....

[Article contains additional citation context not shown here]

S. Manchanda and D. S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. Morgan Kaufmann Publishers, San Mateo, CA, 1988.

....of set oriented updates (bulk updates) becomes an issue. WF97] present an update language based on deferred updates which solves this problem. Other well known approaches in the deductive database community subsumable under the updates in the body paradigm are the early works on DLP [MW88] based on dynamic logic) and LDL updates [NT89] In contrast, frameworks with semantics similar to production rules typically express updates in the head of rules (cf. Sections 2.2 and 4) A main difference between update languages and active rules is that in the former, updates are initiated ....

....be extended to model nested transactions. With nested transactions, Statelog provides a unified framework for modeling several advanced concepts in active databases, e.g. sophisticated coupling modes, event consumption policies, and trigger firing policies. The following example, adapted from [MW88,Che95] motivates why structuring capabilities and a refined transaction model may be useful: Example 4 (To Hire or Not to Hire) Consider relation emp from Example 2. We want to hire an employee only if the average salary after the update does not exceed a certain limit. Such a ....

S. Manchanda and D. S. Warren. A Logic-Based Language for Database Updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pp. 363--394. Morgan-Kaufmann, Los Altos, CA, 1988.

.... 2 interpretation and transformations between belief sets correspond to simple insertions and removals (we are not worried about the permissibility of these operations in a given state) The semantics from the Section 5 is very similar to the formal semantics of Dynamic Logic Programming, see [6]. The dynamic meaning function of [6] assigns to each dynamic predicate symbol a function from ground terms to the binary relations on a set of states (or, equivantely, a function that assigns a binary relation on a set of states to each ground proposition) A accessiblity of the Section 5 assigns ....

.... between belief sets correspond to simple insertions and removals (we are not worried about the permissibility of these operations in a given state) The semantics from the Section 5 is very similar to the formal semantics of Dynamic Logic Programming, see [6] The dynamic meaning function of [6] assigns to each dynamic predicate symbol a function from ground terms to the binary relations on a set of states (or, equivantely, a function that assigns a binary relation on a set of states to each ground proposition) A accessiblity of the Section 5 assigns to propositions functions from ....

S. Manchanda, D.S. Warren, A Logic-based Language for Database Updates, in Foundations of Deductive Databases and Logic Programming, J. Minker (ed.), Morgan Kaufmann Publishers, 1988

....can be performed at definition time to execute update requests more efficiently. 1 Introduction The theory of database updates is currently receiving much more attention than in the past [1] and recently, the problem has been largely studied by various authors with regard to deductive databases [7, 12, 19, 23, 25, 29, 32] and related frameworks [13, 35] This study is particularly important since the main goal of deductive database research is to integrate logic programming and databases in order to provide future database systems with more powerful mechanisms for data representation and computation [8] and sofar ....

....use only the rules and we show that a computation rule suffices and can be freely chosen. Finally, the concept of completeness of updates is defined with respect to operational issues, we give instead completeness results with respect to the declarative semantics. Several other interesting papers [7, 16, 17, 23, 25] deal with issues related to the problem addressed in this paper, but they present some differences that do not allow a quantitative comparison. In [7] an intensional update is considered as a general 36 property, defined by means of a formula involving quantifiers and meta predicates, specifying ....

[Article contains additional citation context not shown here]

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394, Morgan Kauffman, Los Altos, 1988.

....perfect model semantics. 8 Consider the ordering on state terms (2) in section 4.2 13 The Datalog extensions proposed in [2] have a procedural semantics which is defined by a bottom up evaluation and which does not coincide with model based semantics known from the area of logic programming. [14, 15] use dynamic logic to define semantics of updates but also rely on a certain evaluation strategy (top down) 4] propose Transaction Logic, a very general and powerful framework for numerous phenomenons of state change including database updates and planning and consequently leave the standard ....

S. Manchanda and D. S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. Morgan-Kaufmann, Los Altos, CA, 1988.

....at [S] terminates when there are no changes to successive states, ie when [S final ]D = S final 1]D. 2 Transaction Oriented Hierarchical Structuring The need for structuring capabilities and a more elaborated transaction model can be exemplified as follows (this example is adopted from [MW88, Che95] Example1 (To Hire or Not to Hire) The employee Emp with salary Sal should be hired for department Dept provided the average salary after the update does not exceed a certain limit. This may be expressed in flat Statelog (or similarly in the related XY Datalog approach [Zan93] as ....

....It allows to combine deltas and to express hypothetical updates, however no logical semantics is given. Related to our work are approaches dealing with updates in deductive databases. Often, the rule semantics depends on a certain evaluation strategy, e.g. Abi88, AV91, SK96] bottom up) or [MW88, Che95] top down) whereas e.g. MBM95] is like Statelog independent of a certain strategy. However, these works do not cover the ECA rule paradigm of active databases or the 18 concept of nested transactions. Although Statelog allows a very intuitive bottom up reading of rules (cf. ....

[Article contains additional citation context not shown here]

S. Manchanda and D. S. Warren. A logic-based language for database updates. In J. Minker, ed., Foundations of Deductive Databases and Logic Programming, pp. 363--394. 1988.

....can be determined from the existing equations. Seminaive evaluation falls out as a special case when all rules specify inserts. The method is amenable to parallelization. 1 Introduction Database computations with explicit delete and insert operations are gaining more and more attention (e.g. [AV90, AV91a, NK88, MW88]) Such computations arise not only in databases with so called active features (e.g. Esw76, BC79, SHP88, dMS88, MD89, Han89, HJ91, Wid91] but also in production rule systems (e.g. OPS5 [BFKM85] RDL1 [dMS88] However, little is known about the optimization of such computations. The main ....

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. Morgan-Kaufmann, 1988.

....identi er. To get a better grasp of the idea of an oracle as database state, Section 1.4.3 provides some typical examples. 1.4.2 State Transition Oracles The next step is to specify elementary changes to the database. One way to de ne such changes is to build them into the semantics, as in [MW88; NK88; Bon97b; Bon97a; Che91; AV90; McC83] The problem with this approach is that adding new kinds of elementary transitions requires rede ning the very notion of a model and, hence, entails a revamping of the entire theory, including the need to reprove soundness and completeness results. In ....

....database oriented at all, but come from the tradition of procedural programming languages. Typically, they assume that a state is a set of program variables, and that an update changes a variable s value. 1.7.1 Declarative Languages for Database Transactions Dynamic Prolog. Manchanda and Warren [MW88] developed Dynamic Prolog, a logic programming language for database transactions. This language is by far the most similar to T R. For instance, T R and Dynamic Prolog are the only logic programming languages that account not just for updates, but for transaction abort and rollback as well. ....

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363-394. MorganKaufmann, Los Altos, CA, 1988.

....of the new target applications. 8 To conclude this section, note that there is little theory on database evolution. One can note works in a variety of distinct fields: e.g. fixpoint logics (e.g. 6, 7] production rule systems (e.g. 25] or attempts to introduce updates in logic programming [42]. We believe that known formal tools such as temporal logic or more recent ones such as the Evolving Algebra of Gurevich could be used for specifying data evolution. 3.5 Knowledge independence Even if we agree on the importance of data evolution, one may wonder why insist on using rules in place ....

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. MorganKaufmann, 1988.

....they are not always easily used in programming update transaction. On the other hand, in the framework of relational databases, a number of update languages have been defined based on a procedure call oriented execution model, which seems to be easily used in programming transactions on databases [24, 25]. These considerations motivated this study focusing on the possibility of designing an update language for complex objects based on a procedure call oriented execution model. The starting point of this research has been the definition of a suitable set of operators forming the basis upon which ....

....based execution model. Thus, we shall compare our language against two groups of languages: 1. languages for objects with identity, exploiting a bottom up execution model [5, 14, 19] 2. update languages with procedure call based execution models, but defined on the relational model [24, 23, 25]. We start with the languages belonging to the first group. IQL [5] is a strongly typed language for complex object databases based on a bottom up execution model. IQL represents probably the most influential proposal in the context of bottom up languages for manipulating complex data. The data ....

[Article contains additional citation context not shown here]

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394, Morgan Kauffman, Los Altos, 1988.

....also supports sequential updates in which the user can specify the order of updates. A sequential transaction has the form of C 1 ; Cn , which speci es that C i is performed before C j for 1 i j n. The following is an update rule that involves sequential updates. It is adopted from [20]. raise check(E; D;R) raise(E; D;R) avg sal(D; AvgS) AvgS 50K The intention of the rule is to raise the employee salary by R only if the average salary in the department stays below 50K after the raise. This rule is not expressible in U Datalog. The semantics of sequential updates is ....

....into the implementations of ROL [14, 15] ROL2 [17] and Relationlog [16] There are several important issues which we would like to address. Being a database oriented language, DatalogU does not directly support non determinism updates that are important in logic programming based approaches [20, 5, 9, 27]. In LDL [24] the choice operator is used to provide non determinism in deductive databases. We would like to investigate if the explicit use of the existential quanti cation provides a natural way to support non determinism in deductive databases. Besides, we intend to extend DatalogU to support ....

S. Manchanda and D. S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundation of Deductive Databases and Logic Programming, pages 363-394. Morgan Kaufmann Publishers, 1988.

.... databases with deductive capabilities [1, 4, 5, 12] at integrating production rules (traditionally used in expert systems) in database environments [11, 15, 16, 17, 20] and finally at designing update languages better suited than the traditional ones for these new and advanced frameworks [2, 13, 10]. All these efforts have already pro This work was partially supported by Consiglio Nazionale delle Ricerche. y Permanent affiliation: DEIS, Universit a della Calabria, 87036 Rende, Italy. luigi deis21.deis.unical.it. z Part of this work was done while this author was visiting the UCLA ....

S. Manchanda and D.S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394, Morgan Kauffman, Los Altos, 1988.

....and update propagation. We model the sources of information as deductive databases, sharing the same logical language to express queries and updates, but containing independent, even if possibly related, data. Several approaches have been proposed integrating logic and updates (see for example [1,2,6,7,13,14,32,34,35,38,45]) The language we propose is based on the U Datalog language [6] and extends it with support for active rules and with the ability to model heterogeneous databases. U Datalog has been introduced with the aim of providing a set oriented logical update language, guaranteeing update parallelism in ....

S. Manchanda and D.S. Warren. A Logic-Based Language for Database Updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363-394. Morgan-Kaufmann, 1988.

....In particular, updates can be performed as soon as they are generated, as side e ect of the query evaluation, thus, by applying an immediate semantics. Languages based on an immediate semantics include LDL (Naqvi Tsur, 1989) TL (Abiteboul Vianu, 1991) DL (Abiteboul Vianu, 1991) DLP (Manchanda Warren, 1988), Statelog (Lausen et al. 1998) The immediate semantics is in contrast with the deferred update semantics, by which updates are not applied as soon as they are generated during the query evaluation; rather, they are executed only when the query evaluation is completed. Languages based on a ....

....operational mechanism handling negated atoms in U Datalog programs, providing a marking phase and a compositional semantics. The proposed extension is based on the notion of strati cation, rst proposed for logic programming and deductive databases (Abiteboul Vianu, 1991; Lausen et al. 1998; Manchanda Warren, 1988; Naqvi Tsur, 1989) This extension is not, however, a straightforward extension of previously de ned strati cation based semantics for two main reasons. First of all, U Datalog rules are not range restricted (Ceri et al. 1990) but are required to be safe through query invocation, 1 resulting ....

Manchanda, S., & Warren, D.S. (1988). A Logic-Based Language for Database Updates. Pages 363-394 of: Minker, J. (ed), Foundations of Deductive Databases and Logic Programming. Plenum Press, New York.

....a notion of object identifier is modeled by way of existential quantification. As far as the dynamic aspects are concerned, there are proposals which provide a procedural semantics, as in the case of [Mon93, BGM95] and other proposals provide adhoc model theoretic semantics related to modal logic [BK93, MW86, Man89, Che91]. In the former case, a poor integration of the various aspects is achieved, whereas in the latter a close relationship to real computing is missing. Moreover, it is difficult to combine the structural and dynamic aspects of the object data model with deduction. For instance, the F logic combined ....

S. Manchanda and D. S. Warren. A logic-Based Language for Database Updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. SpringerVerlag, Berlin, 1986.

....hypothetical rules, in which the premise can query not only a real database, but hypothetical databases as well. Vieille, et al., for instance, have developed a deductive database along these lines [66, 67] and Warren and Manchanda have used hypothetical rules to reason about database updates [68, 44]. In [50] Miller shows that hypothetical insertions can structure the runtime environment of logic programs, resulting in programs that are more elegant, more efficient, and easier to maintain. In [51] he develops a theory of lexical scoping based on the hypothetical creation of constant symbols ....

....6.4 that R; DB j= OE if and only if MR ; DB j= OE. The result then follows by Lemma 6.2. 2 7. RELATED WORK There has been considerable work in the area of logic and updates. The logical systems that are the closest to Hypothetical Datalog are Dynamic Prolog, developed by Manchanda and Warren [45, 44, 68], and EKS, developed by Vieille, et al. [66, 67] All three systems deal with both insertion and deletion, and in all three, updates take place on base facts, not derived facts. Nevertheless, the three systems are very different. In EKS, the semantics is purely operational. The emphasis here has ....

S. Manchanda and D.S. Warren. A Logic-based Language for Database Updates. In Jack Minker, editor, Foundations of Deductive Databases and Logic Programming, chapter 10, pages 363--394. Morgan Kaufmann, 1988.

....for the representation of updates in which the sequential composition of updates is handled explicitly and arbitrary basic updates can be integrated. Transaction Logic has no explicit construct for bulk updates. Of course, the e ect of bulk updates can be obtained using recursive rules (see e.g. MW88b] for an example) In [BKC93] a relational assignment operator for copying an IDB relation into the EDB is proposed. However, this operation is atomic and outside the scope of Transaction Logic. Thus, bulk updates cannot be composed from existing single updates. The general concept of Transaction ....

....always gets a bulk update e ect and cannot specify update alternatives as in Transaction Logic and ULTRA. In addition, the aggregation of the update requests does not distinguish between logical conjunction and disjunction. The expressibility of the language DLP proposed by Manchanda and Warren [MW88b] is comparable to that of the rule fragment of Transaction Logic restricted to basic insertions and deletions. Like in our approach, the implementation uses a non immediate update semantics. The formal semantics, however, is based on a dynamic logic in which a formula is interpreted by pairs of ....

[Article contains additional citation context not shown here]

S. Manchanda and D. S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363-394. Morgan Kaufmann Publishers, San Mateo, CA, 1988.

No context found.

S. Manchanda and D. S. Warren. A logic-based language for database updates. In J. Minker, editor, Foundations of Deductive Databases and Logic Programming, pages 363--394. Morgan Kaufmann Publishers, San Mateo, CA, 1988.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC