J. Kiernan, C. de Maindreville, and E. Simon. Making deductive databases a practical technology: a step forward. In ACM SIGMOD Int. Conf. on Management of Data, pages 237--246, 1990.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....unjustified tuple remains in non tc. 2 Although the given program may be rewritten using a (somewhat intricate) technique for delaying rules, a better solution is to use one of the declarative semantics developed for logic programs whenever the use of negation is important; see Section 2.3. RDL1 [KdMS90] is a deductive database language with production rule semantics; a rule algebra is used as an additional control mechanism. A RDL [SK96] extends RDL1 by active database concepts, in particular delta relations and a module concept. 2.3 Deductive Rules The logic programming and deductive ....

G. Kiernan, C. de Maindreville, and E. Simon. Making Deductive Database a Practical Technology: a step forward. In ACM Intl. Conference on Management of Data (SIGMOD), pp. 237--246, 1990.

....(and production) rules into a system that supports only event driven rules by translating each high level rule into one event driven rule. The translation is based on the emulation of a deductive rule through a special if then else production rule. This method is also used in the RDL1 system [17], under a more user controlled and low level deductive rule scheme. Production rules in DEVICE are integrated into the active OODB using the same scheme. The conditions of both high level rule types are incrementally matched using a complex event network that resembles the RETE like discrimination ....

....in the head of the rule exists in the database. This is a declarative way to state that when the condition is satisfied, then the derived object should be inserted in the database. According to this procedural interpretation, deductive rules can be emulated using production rules, as in RDL 1 [17]. However, the simple production rule translation scheme must be extended to fully capture the semantics of deductive rules. For example, the creation of a new derived object should only be done if the object does not already exist, otherwise two distinct objects with the same attribute values ....

J. Kiernan, C. de Maindreville and E. Simon, Making deductive databases a practical technology: A step forward, Proc. ACM SIGMOD Int. Conf. on the Management of Data, (1990) 237246.

....the join with a virtual node is done by querying directly the original objects. Equivalently, the deletion of a token from a virtual node is not burdened by any look up costs at the two input event memories. Finally, DEVICE has been extended with deductive rules as well, based on the assumption [27] that the head of a deductive rule is equivalent to the insertion of a new derived object in the database, while the condition matching is equivalent to production rules. What had to be altered was a) the rule control strategy in order to consider stratification, and b) the semantics of deletion ....

Kiernan, J, de Maindreville, C and Simon, E, Making deductive databases a practical technology: A step forward, in: Garcia-Mollina, H and Jagadish, H V, eds., Proc. ACM SIGMOD Int. Conf. on the Management of Data, 1990, 237-246.

....is built from scratch. ECA rules are emulated using production rules and special differential files (or delta relations) that record data modifications to compensate for the lack of pure events in a relational database environment. The same architectural line is followed in the RDL1 A RDL system [24, 35] which additionally supports deductive rules by treating them as special production rules (see section 5.3) A rather different approach is followed by U Datalog [7, 6] where deductive rules are extended with updates; therefore, production rules can be emulated. U Datalog allows updates to appear ....

....path(start:X,finish:Y) A arc(start:Z,finish:Y) THEN path(start:X,finish:Y) 13 When a new arc object is created then the first rule is executed and a new path object is also created. According to this procedural interpretation, deductive rules can be emulated using production rules, as in RDL1 [24]. However, the simple production rule translation scheme must be extended to fully capture the semantics of deductive rules. For example, the creation of a new derived object should only be done if the object does not already exist, otherwise two distinct objects with the same attribute values ....

J. Kiernan, C.d. Maindreville, and E. Simon, "Making deductive databases a practical technology: A step forward," Proc. ACM SIGMOD Int. Conf. on the Management of Data, Atlantic City, NJ, 1990, pp. 237-246.

....some systems support or emulate deductive rules as well. For example, Ariel [28] supports both implicit and explicit triggers under a data driven rule scheme, while OPRA [35] a recent extension to the ADAM OODB, emulates implicit triggers into an event driven rule environment. The 3 RDL1 system [31] supports deductive rules in a production rule environment, while it provides active rule extensions (A RDL) for specifying and detecting events in the deductive rule language [40] Finally, POSTGRES [38] and Chimera [8] support backward chaining deductive rules through data retrieval events, ....

.... attempts to integrate deductive and active databases [30, 40, 45] The RDL1 ARDL systems integrate all three kinds of rules, by supporting event specifications in the deductive rule language [40] through delta relations, while deductive rules are emulated using production rules at the lower level [31]. OPRA emulates production rules in the event based active OODB ADAM [12] using several ECA rules. This approach is also followed in [9] for translating deductive rules into the active relational database system Starburst [44] where originally only event driven rules were supplied. Fig. 1 ....

[Article contains additional citation context not shown here]

J. Kiernan, C. de Maindreville and E. Simon, Making deductive databases a practical technology: A step forward, Proc. ACM SIGMOD Int. Conf. on the Management of Data, Atlantic City, NJ (1990) 237-246.

....languages and, more recently, as high level protocols to interact with heterogeneous data repositories, is the most remarkable success of declarative languages in this field. The natural evolution of this research has led to the use of logic based languages, like Datalog [1, 28] LDL [34] or RDL [17], that add deductive power to the simple databases we were used to work with. These deductive databases can efficiently handle vast amounts of derived knowledge, moving most semantic issues from (diffuse) applicative logic to (centralized) database design. The second main goal has been to make ....

J. Kiernan, C. de Maindreville, and E. Simon. Making deductive databases a practical technology: A step forward. In Proc. of the ACM SIGMOD International Conference on the Management of Data, pages 237--246, Atlantic City, New Jersey, May 1990. 27

....within a program. However, we would like to offer trigger or demon like implicit activation as well depending on declaratively specified conditions. Again we are looking for a clean way of syntactically integrating implicit activation into the overall framework of the language. Solutions like [KMS90] exclusively based on a production rule like setting are not acceptable for us. 3.4 Towards a Flexible Integration We believe that the combination of features which KBPL is offering is particularly well suited for achieving the goal of a flexible integration of declarative and procedural style ....

G. Kiernan, C. de Maindreville, and E. Simon: "Making Deductive Databases a Practical Technology: A Step Forward", in: Proc. 16th ACMSIGMOD Conf. on Management of Data, Atlantic City/NJ, 1990, 237-246

....research activity for active DBMSs, prior to this paper the important problem of completely integrating declarative constraints and active rules without restrictions has not been addressed. The execution models of many prototype active database systems, e.g. Ariel [14] HiPac [8, 18] RDL [22], RPL [9] and Starburst [36] consider only ECA rules and are more in the style of OPS5; the activation of triggers is considered only at preset activation points, the execution of trigger actions are non interruptible, and trigger activation is determined based on the net effect of multiple ....

J. Kiernan, C. de Maindreville, and E. Simon. Making deductive databases a practical technology: A step forward. In Proc. ACM SIGMOD Int. Conf. on Management of Data, pages 237--246, Atlantic City, New Jersey, June 1990.

....when using priorities, it is not possible to specify different control structures for rules triggered by different events. Additionally, since the control structures of rule execution are implicitly specified by rule priorities, it is quite difficult to understand or modify the rule structures. In [26, 39], Simon et al. emphasize the need for a more powerful rule control mechanism for making a rule system more practical. They propose a set of control constructs, namely sequence , disjunction and saturate , for specifying the control structures among rules. In RDL1 (the active DBMS proposed by ....

G. Kiernan, C. deMaindreville, and E. Simon. Making deductive database a practical technology: a step forward. In Proc. ACM SIGMOD Int'l Conf. on Management of Data, 1990.

....temporal aggregate functions. The importance of such functions was observed by the temporal database researchers, e.g. 33] Other work on Triggers in Active Databases: The other relevant work is the work on active databases (e.g. 6, 8, 5, 20] on triggers [4, 12, 26] and on rule languages [22, 27, 29, 42, 21]. These works lack the temporal component in the following sense. They concentrate on the specification of triggers that involve at most two database states: the current one, and the previous one. In contrast, in this paper we address the more general temporal aspects in triggers. Also, the ....

G. Kiernan, C. de Maindreville, E. Simon, Making Deductive Database a Practical Technology: A Step Forward, Proc. of the ACM-Sigmod International Conf. on Management of Data, 1990.

....then the action of the rule is executed. Different aspects of support for ECA rules are presented in [32] Active database systems, however, have much in common with production systems (which support condition action rules) and some active databases provide direct support for production rules [27, 25]. For certain tasks which monitor the state of the database, such as integrity constraint checking or maintenance of derived data, many ECA rules may be required to express the same functionality as a single production rule. Thus systems which support only ECA rules may provide cumbersome ....

G. Kiernan, C. de Maindreville, and E. Simon. Making Deductive Databases a Practical Technology: a step forward. In H. Garcia-Molina and H.V. Jagadish, editors, Proc. ACM SIGMOD Conf, pages 237--246, 1990.

....techniques that are also important to LDL include the unification of complex terms, the compile time analysis to detect unsafe queries, and the compiling of safe queries. These are discussed in [165, 166] 2.2. 7 RDL1 and RDL C Unlike LDL [108, 154] or other deductive database systems [155] RDL1 [32, 76] rules are not in clausal form in the style of PROLOG or DATALOG, but closer to forward chaining rule based languages like OPS5. Again, 31 the design goal is to integrate production rule language with a relational DBMS. A rule consists of a condition part which is any tuple relational calculus ....

....is of crucial importance in the process of decomposition abstraction. In this section, we analyze the concurrent behavior of each benchmark program and point out key issues to the successful application of proposed mechanisms. 7.2. 1 MANNERS MANNERS was derived from an example program in [76] which employed a combinatorial search for solving a seat assignment problem among a number of guests. The seats must be assigned such that neighbors are of opposite sex and share at least one common hobby. This simple program, containing just 8 rules, is a very good test program for evaluating ....

G. Kiernan, C. de Maindreville, and E. Simon. Making deductive database a practical technology: A step forward. In ACM SIGMOD Intl. Conf. on Management of Data, pages 237--246, 1990.

....intermediate facts this is important with bottom up computation, since facts that are computed are generally not discarded anywhere else, and would use excessive amounts of memory. Modules with run time semantics are also available in several production rule systems (for example, RDL1 [KdMS90] 9 Future Directions A number of issues require further work. These include support for metaprogramming, constraints, diskresident data, new data types and operations, user interfaces, inheritance and object orientation. 10 Acknowledgements We would like to acknowledge our debt to LDL, ....

G. Kiernan, C. de Maindreville, and E. Simon. Making deductive database a practical technology: a step forward. In Proceedings of the ACM SIGMOD Conf. on Management of Data, 1990.

....claim that although some database rule languages are clearly classifiable as either deductive or active, there is no obvious delineation between the two types of languages. We support our claim by considering six relational database rule languages prominent in the literature: Datalog [Ull89] RDL [KdMS90], A RDL [SKdM92] Ariel [Han89, Han92] Starburst [WCL91, WF90] and Postgres [SJGP90] We believe that these six rule languages form a spectrum along a deductive to active axis , as shown in Figure 1. We also believe that the same spectrum describes the level of abstraction of these database ....

J. Kiernan, C. de Maindreville, and E. Simon. Making deductive databases a practical technology: A step forward. In Proceedings of the ACM SIGMOD International Conference on Management of Data, pages 237--246, Atlantic City, New Jersey, May 1990.

....it is not readily apparent how to do so. Our work also addresses systems issues such as transactions, no update triggers, etc. Other work on Triggers in Active Databases: The other relevant work is the work on active databases (e.g. 8, 9, 7, 18] on triggers [6, 11, 26] and on rule languages [21, 28, 30, 39, 20]. These works lack the temporal component in the following sense. They concentrate on the specification of triggers that involve two database states: the current one, and the previous one. In contrast, in this paper we address the more general temporal aspects in triggers. Temporal Databases: The ....

G. Kiernan, C. de Maindreville, E. Simon, Making Deductive Database a Practical Technology: A Step Forward, Proc. of the ACM-Sigmod International Conf. on Management of Data, 1990.

....LDL, where modules have only a compile time meaning and no run time meaning, modules in CORAL have important run time semantics. Several run time optimizations are done at the module level. Modules with run time semantics are also available in several production rule systems (for example, RDL1 [KdMS90] LDL [AO93] a successor to LDL, has moved in the direction taken by CORAL in several respects. It is interpreted, supports abstract data types, and uses a local semantics for choice (Carlo Zaniolo, personal communication) XSB [SSW93] is a system being developed at SUNY, Stony Brook. It ....

G. Kiernan, C. de Maindreville, and E. Simon. Making deductive database a practical technology: a step forward. In Proceedings of the ACM SIGMOD Conference on Management of Data, 1990.

....LDL, where modules have only a compile time meaning and no run time meaning, modules in CORAL have important run time semantics, in that several run time optimizations are done at the module level. Modules with run time semantics are also found in several production rule systems (for example, RDL1 [8]) LDL , a successor to LDL under development at MCC Austin, is reportedly also moving in the direction taken by CORAL in many respects. It will be partially interpreted, support abstract data types, and use a local semantics for choice (Carlo Zaniolo, personal communication) XSB is a system ....

G. Kiernan, C. de Maindreville, and E. Simon. Making deductive database a practical technology: a step forward. In Proceedings of the ACM SIGMOD Conference on Management of Data, 1990.

....a framework for translating constraint specifications into production rules that maintain the constraints. Production rules in database systems allow specification of data manipulation operations that are automatically executed whenever certain events occur and or certain conditions are met [DE89, Han89,KdMS90,SJGP90,WF90]. The usefulness of incorporating production rules into database systems is well accepted [EC75,MD89,Mor83] particularly in the context of constraint enforcement. However, we know of no automatic (or semi automatic) method for specifying general constraints in a high level, non procedural ....

J. Kiernan, C. de Maindreville, and E. Simon. Making deductive databases a practical technology: A step forward. In Proc. of ACM-SIGMOD, pages 237--246, May 1990.

....of RPL incorporate a programmable conflict resolution mechanism, imparting a high degree of flexibility to the language. However, problems arising 42 from database scaling are not addressed in the RPL system by providing support for balanced parallel rule processing. ffl RDL1: The RDL1 system [44] includes features found in expert system shells and supports two input languages, RDL1 and an extended version of SQL. The underlying data model is the relational data model augmented with abstract data types (ADTs) The ADT capability is supported by user defined data types and user defined ....

....provides a declarative framework for conflict resolution like that proposed for PARULEL. However, it appears that the attributes of the firing groups used for conflict resolution are hard coded; there is no actual way of accessing the conflict set itself. Another system related to PARULEL is RDL [44]. Although the execution semantics is defined as non deterministic choice for rule execution, control can be specified separately from the rules by way of a control language based upon regular expressions. In early work [74] one of the authors defined a rule language controlled by a separate ....

[Article contains additional citation context not shown here]

G. Kiernan, C. de Maindreville, and E. Simon. Making deductive database a practical technology: a step forward. In Proc. of the ACM-SIGMOD 1990, Intl. Conf. on the Management of Data, pages 237--246, 1990.

....memoing top down evaluation used in XSB [War89] This is a tuple at a time approach. Future extensions supporting constraints in tuples are planned. We note that there are other projects, such as the RDL effort at INRIA, that have not focused upon deductive databases, but are closely related [KdMS90] 6 Deductive Database System Implementations There have been a number of implementations of deductive databases. The results of a survey conducted by the authors over the Internet, presented below, indicate that the extensive published research on deductive databases was accompanied by quite a ....

G. Kiernan, C. de Maindreville, and E. Simon. Making deductive database a practical technology: a step forward. In Proceedings of the ACM SIGMOD Conference on Management of Data, 1990.

....The server has been implemented and tested on a shared nothing multiprocessor architecture. 1 Introduction Databases coupled with active rules, which are known as active databases, are becoming increasingly popular because of their added features for supporting a wide spectrum of applications [27, 47, 48, 23, 8, 29, 50, 3, 14, 18, 24, 44]. In contrast to the traditional database systems, active systems monitor a variety of events that can occur and react to them automatically. In the majority of these systems, the active behavior is incorporated using production rules or simply rules. Although, there have been several efforts ....

.... the operation and send the operation to data processor; Until end of the task; Send a signal to GTS signal queue that the task is completed; Forever; End LTM; 5 Related Work There have been several research efforts on execution models for the rules and transactions of active database systems [27, 37, 47, 8, 23, 48, 46, 41, 29, 50, 5, 18, 38, 3, 24, 6, 26, 16]. However, very few of them focus on the design and implementation issues in shared nothing parallel environments. Although our system as an integrated parallel active database server is unique, we shall compare our parallel rule execution model and the transaction model with that of some related ....

G. Kiernan, C. deMaindreville, and E. Simon. Making deductive database a practical technology: a step forward. In Proc. ACM SIGMOD Int'l Conf. on Management of Data, 1990.

....deductive databases can be extended to support active behavior, e.g [BJ93, CCCR 90, HD93b, SKdM92,Tan91] this relates to the converse of the problem considered here. In the RDL1 system, a forward chaining production style of rule processing is used to implement a deductive database language [KdMS90] The goal of RDL1 is an efficient deductive database system. Hence, although the underlying implementation is based on rule triggering, the rule language is deductive. In our case, we illustrate how deductive rules can be supported using the existing facilities of an active database system. This ....

J. Kiernan, C. de Maindreville, and E. Simon. Making deductive databases a practical technology: A step forward. In Proceedings of the ACM SIGMOD International Conference on Management of Data, pages 237--246, Atlantic City, New Jersey, May 1990.

No context found.

J. Kiernan, C. de Maindreville, and E. Simon. Making deductive databases a practical technology: a step forward. In ACM SIGMOD Int. Conf. on Management of Data, pages 237--246, 1990.

No context found.

J. Kiernan, C. de Maindreville, and E. Simon. Making Deductive Databases a Practical Technology: A Step Forward. Proc. of the ACM SIGMOD Int'l Conf. on Management of Data, pp. 237-246, Atlantic City, NJ, May 1990.

No context found.

Kiernan, G., de Maindreville, C., Simon, E. Making deductive databases a practical technology: a step forward. INRIA Research Report 1153, 1990.

First 50 documents

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