R. Ramakrishnan, D.Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the Int'l VLDB Conf., Vancouver, Aug. 1992.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....of integer, multiset of real, multiset of string) User defined types may be defined over existing types using a tuple constructor. The instances of a user defined type are uniquely identified by an object identifier (OID) The RODIN system design uses the CORAL Deductive Database System [33] as an implementation engine. The CORAL system is used to: 1) store the system catalog; 2) store (a representation of) schemas that have been imported from external systems or generated as a result of integrating existing schemas; and 3) store information about external sites and DBMS s that can ....

....to create a nested set instead of multiset) Corresponding functions that compute the aggregate on the set of distinct values in each multiset are also supported, for example avg distinct, sum distinct, prod distinct, or count distinct. A detailed description of the CORAL system is available [33]. Additional CORAL facilities that are used in the design and implementation of RODIN are described below. CORAL programs are used to represent the mapping informationneededinquerytranslationinRODIN. Foreach construct in a schema that is derived in the importation, restructuring, or integration ....

R. Ramakrishnan, D.Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the Int'l VLDB Conf., Vancouver, Aug. 1992.

....checking. Indeed, for rangerestricted programs, they have been proven to be asymptotically equivalent [10, 8] under certain assumptions. Despite these well known equivalences, magic style systems have traditionally differed from tabling systems. Magic style systems, such as Aditi [15] CORAL [7], and LDL [3] are built upon set at a time semi naive engines, while tabling systems, such as XSB [9] use a tuple at a time strategy that reflects their genesis in the logic programming community. Each class of systems has its advantages and disadvantages. Presently for in memory Datalog ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proceedings of the 18th VLDB, pages 238--250, 1992.

....with bottom up redundancy checking. In fact, under certain assumptions they have been proved to be asymptotically equivalent [Sek89] Despite these well known equivalences, magic style systems have traditionally differed from tabling systems. Magic style systems, such as LDL [CGK 90] CORAL [RSS92b] Glue Nail [DMP93] and Aditi [VRK 94] are built upon set at a time engines, and can use set at a time operations like relational joins that may be made efficient for disk resident data, while tabling systems, such as XSB [SSW94] use a tuple at a time strategy that reflects their genesis ....

....problems, but available tabling systems such as XSB [SSW 97] still used a tuple at a time strategy. Even though caching techniques may alleviate the problem for non recursive queries, for recursive queries it is not known in advance how much should be cached. Bottom up systems such as Coral [RSS92b] Aditi [VRK 94] and LDL [CGK 90] can evaluate recursive queries in a set at a time fashion. However, the experimental results of [SW94b, SSW94] have shown that for in memory queries, these systems perform rather poorly compared to XSB. A number of factors contribute to the better ....

[Article contains additional citation context not shown here]

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proceedings of VLDB, pages 238--250, 1992.

....checking. Indeed, for range restricted programs, they have been proven to be asymptotically equivalent [15, 13] under certain assumptions. Despite these well known equivalences, magic style systems have traditionally differed from tabling systems. Magic style systems, such as LDL [4] CORAL [12], Glue Nail [5] are built upon set at a time semi naive engines, while tabling systems, such as XSB [14] use a tuple at a time strategy that reflects their genesis in the logic programming community. Each class of systems has its advantages and disadvantages. Presently for in memory Datalog ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

....checking. Indeed, for range restricted programs, they have been proven to be asymptotically equivalent [13, 11] under certain assumptions. Despite these well known equivalences, magic style systems have traditionally differed from tabling systems. Magic style systems, such as LDL [4] CORAL [10], Glue Nail [5] are built upon set at a time semi naive engines, while tabling systems, such as XSB [12] use a tuple at a time strategy that reflects their genesis in the logic programming community. Each class of systems has its advantages and disadvantages. Presently for in memory Datalog ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

....of a top down evaluation in a bottom up framework. Nevertheless, producing an efficient bottom up engine, that moreover can be smoothly integrated with the procedural power of Prolog is a difficult task that has not yet been completely accomplished, despite such noteworthy efforts as CORAL [8], LDL [4] LDL [17] GLUE NAIL [5] ADITI [18] EKS V1 [19] among others. For in memory deductive database queries, SLG resolution [3, 2] offers an alternative implementation strategy to those based on bottom up evaluation of rewritten programs. SLG is a top down, tuple at a time tabling method ....

R. Ramakirshnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

....general fixed rules throughout the paper. All results can be easily extended to any fixed computation rule. so that it is modularly stratified for a fixed computation rule. This property has made modular stratification useful for deductive database systems based on magic set evaluation (e.g. [11]) However, whether a program is modularly stratifiable at all is undecidable. Furthermore, programs which can be evaluated using a fixed computation rule may not be modularly stratified, as the program in the following example. Example 2. The following program is neither modularly, nor weakly ....

....semantics of all normal programs. Despite the fact that our results have been presented using (variants of) SLG resolution, we believe that the underlying ideas are potentially applicable to deductive database systems whose engines have radically different designs from XSB. For example, CORAL [11] evaluates modularly stratified programs using Ordered Search [10] Described simply, this technique keeps a tree of dependencies and restricts magic evaluation to magic facts that are in an SCC. Both Ordered Search and its extension Well Founded Ordered Search as originally formulated do not ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL --- Control, Relations, and Logic. In Proceedings of the 18th Conference on Very Large Data Bases, pages 238--249, Vancouver, Canada, August 1992.

.... negation and recursion in datalog has proven to be a major issue, and the deductive database community has made significant contributions to developing the necessary underlying semantics [21] Further, prototype deductive database systems, exemplified by LDL [6] Glue Nail [7] and CORAL [12], have been designed and developed to incorporate these advances. It should be mentioned that all these systems process all recursive data in memory. XSB offers an alternative approach to creating a deductive database system. Rather than depending on rewriting techniques, it extends Prolog s SLD ....

....publicly available, the comparisons that are provided below are made using published times for certain queries. Aditi is a multi user system which allows direct computation on disk resident data, and so comparisons with a single user system for memory resident queries seem pointless. As for LDL, [12] presents a comparison of CORAL with LDL, and indicates that for most simple queries, CORAL is usually significantly faster than LDL (an assessment largely echoed in [8] As a result, the comparisons in this section are primarily made against against CORAL, although we include supporting ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

....for our custom notations for high level design. Figure 1 shows the architecture of our prototype environment. The environment features visual editors for formulating both software designs (as labelled rectangles and edges) and ISF interconnection styles. Central to the architecture is the Coral [13] deductive database system. Visual software designs are translated into Datalog facts and interconnection styles are translated into Datalog rules. Using Coral, designers can verify the well formedness of particular designs. The verification process will detect stylistic violations. Examples of ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, Relations, and Logic. In Proceedings of the International Conference on Very Large Data Bases, pages 238--250, 1992.

....of this class of programs, is that if a program is modularly stratified, then it is statically reorderable so that it is modularly stratified for a fixed computation rule. This property has made modular stratification useful for deductive database systems based on magic set evaluation (e.g. [21]) However, whether a program is modularly stratifiable at all is undecidable. Furthermore, programs which can be evaluated using a fixed computation rule may not be modularly stratified, as shown by the program in the following example. Example 1.2 The following program is neither modularly, nor ....

....system mentioned in the introduction. Despite the fact that our results have been presented using (variants of) SLG resolution, we believe that the underlying ideas are potentially applicable to deductive database systems whose engines have radically different designs from XSB. For example, CORAL [21] evaluates modularly stratified programs using Ordered Search [20] Described simply, this technique dynamically builds a tree of dependencies and restricts magic evaluation to magic facts that are in a set of mutually dependent subqueries. Both Ordered Search and its extension Well Founded ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL --- Control, Relations, and Logic. In Proceedings of the 18th Conference on Very Large Data Bases, pages 238--249, Vancouver, Canada, August 1992. Morgan-Kaufmann.

....of fixed computation rules is that if a program is modularly stratified, then it is decidably reorderable so that it is modularly stratified for a fixed computation rule. This property has made modular stratification greatly useful for deductive databases, and programing languages (e.g. [8]) However, whether a program is modularly stratifiable at all is undecidable. Furthermore, programs which can be evaluated in a fixed order may not be modularly stratified, as in Example 1.2. Example 1.2 The following program is neither modularly stratified, nor weakly stratified [5] s: ....

....the SLG WAM for all normal programs, using delay and simplification transformations to break the fixed order search when the engine enters a flummoxed state. However the results in this section are potentially applicable to deductive database engines of radically different designs such as CORAL [8] which evaluates modularly stratified programs using the technique of Ordered Search [7] Described simply, Ordered Search keeps a tree of dependencies and restricts magic evaluation to magic facts that are in an independent SCC. Ordered Search as originally formulated does not contain a mechanism ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

....evaluation in a bottom up framework. Nonetheless, producing an efficient bottom up engine, that moreover can be smoothly integrated with the procedural power of Prolog is a difficult enterprise that we believe has not yet been completely accomplished, despite such noteworthy efforts as CORAL [15], LDL [6] and GLUE NAIL [7] SLG resolution ( 5] 2] offers an alternative implementation strategy to those based on magic evaluation methods. SLG is complete for non floundering programs with finite models, whether they are stratified or not 1 . SLG has been partially implemented by the ....

....For negation, comparisons are made against published times for certain queries to GLUE NAIL also. In the case of Aditi, which is a multi user system allowing direct computation on disk resident data, comparisons with a single user system for memoryresident queries seem pointless. As for LDL, [15] presents a comparison of CORAL with LDL, and indicates that for most simple queries, CORAL is usually significantly faster than LDL (an assessment largely echoed in [10] Further performance analysis can be found in [24] 6.1 Benching Environment A pre release of XSB version 1.3 was benched ....

[Article contains additional citation context not shown here]

R. Ramakirshnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

....is well formed, with respect to an ISF specification, when the configuration of the entities and relations that comprise the design do not violate any of the rules of the ISF specification. Figure 1 shows the architecture of our prototype environment. Central to the architecture is the Coral [6] deductive database system. Software designs are encoded as Datalog [8] facts and interconnection styles are encoded as Datalog (a Prolog like language) programs. Using Coral, designers can verify the wellformedness of particular designs, as well as make queries about a particular design. For ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, Relations, and Logic. In Proceedings of the International Conference on Very Large Data Bases, pages 238--250, 1992.

....evaluation in a bottom up framework. Nonetheless, producing an efficient bottom up engine, that moreover can be smoothly integrated with the procedural power of Prolog is a difficult enterprise that we believe has not yet been completely accomplished, despite such noteworthy efforts as CORAL [14], LDL [6] and GLUE NAIL [7] SLG resolution [5, 4] offers an alternative implementation strategy to those based on magic evaluation methods. SLG is complete for non floundering programs with finite models, whether they are stratified or not 1 . SLG has been implemented in the XSB logic ....

....comparison can be made using published times for certain queries, comparisons which will be provided below. Aditi is a multi user system which allows direct computation on disk resident data, and so comparisons with a single user system for memory resident queries seem pointless. As for LDL, [14] presents a comparison of CORAL with LDL, and indicates that for most simple queries, CORAL is usually significantly faster than LDL (an assessment largely echoed in [9] As a result, the comparisons in this section are usually made against against CORAL, although we include supporting ....

[Article contains additional citation context not shown here]

R. Ramakirshnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

....checking. Indeed, for rangerestricted programs, they have been proven to be asymptotically equivalent [10, 8] under certain assumptions. Despite these well known equivalences, magic style systems have traditionally differed from tabling systems. Magic style systems, such as Aditi [15] CORAL [7], and LDL [3] are built upon set at a time semi naive engines, while tabling systems, such as XSB [9] use a tuple at a time strategy that reflects their genesis in the logic programming community. Each class of systems has its advantages and disadvantages. Presently for in memory Datalog ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proceedings of the 18th VLDB, pages 238--250, 1992.

....a top down evaluation in a bottom up framework. Nevertheless, producing an efficient bottom up engine, that moreover can be smoothly integrated with the procedural power of Prolog is a difficult task that we believe has not yet been completely accomplished, despite such noteworthy efforts as CORAL [8], LDL [4] GLUE NAIL [5] ADITI [17] EKS V1 [18] and others. Supported in part by the National Science Foundation under Grant No. CCR 9102159. For in memory deductive database queries, SLG resolution [3, 2] offers an alternative implementation strategy to those based on magic or other ....

R. Ramakirshnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

.... closed world negation and recursion in datalog has proven to be a major issue, and the deductive database community has made significant contributions to developing the necessary underlying semantics [21] And prototype deductive database systems, exemplified by LDL [6] Glue Nail [7] and CORAL [12], have been designed and developed to incorporate these advances. We should mention that all these systems process all recursive data in memory. XSB offers an alternative approach to creating a deductive database system. Rather than depending on rewriting techniques and a bottom up evaluation ....

....publicly available, the comparisons that are provided below are made using published times for certain queries. Aditi is a multi user system which allows direct computation on disk resident data, and so comparisons with a single user system for memory resident queries seem pointless. As for LDL, [12] presents a comparison of CORAL with LDL, and indicates that for most simple queries, CORAL is usually significantly faster than LDL (an assessment largely echoed in [8] As a result, the comparisons in this section are primarily made against against CORAL, although we include supporting ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proceedings of the 18th International Conference on Very Large Data Bases, pages 238--249. VLDB Endowment, 1992.

....redundancy checking. Indeed, for range restricted programs, they have been proven to be asymptotically equivalent [10, 8] Despite these well known equivalences, magic style systems have traditionally differed from tabling systems. Magic style systems, such as Aditi [12] LDL [2] and CORAL [7], are built upon set at atime engines, and can use set at a time operations like relational joins that may be made efficient for disk resident data, while tabling systems, such as XSB [9] use a tuple at a time strategy that reflects their genesis in the logic programming community. Presently for ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th International Conference on Very Large Data Bases (VLDB), pages 238--250, 1992.

....and bottom up mechanisms to evaluate queries written in a logical language. In magic evaluation, which is set at a time, the top down component is an analysis step which transforms a program so that it can be executed by a bottomup engine in a goal directed manner. Some magic based systems, [8] and [6] also provide a top down execution mode (pipelining) which can interact with bottom up evaluations. Deductive database systems can also be built upon tuple at a time evaluation methods, such as SLG [5] SLG allows resolution of subgoals either by program clauses, providing a top down ....

R. Ramakirshnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proc. of the 18th Int'l Conf. on Very Large Data Bases, pages 238--249. VLDB End., 1992.

....with bottom up redundancy checking. In fact, under certain assumptions they have been proved to be asymptotically equivalent [Sek89] Despite these well known equivalences, magic style systems have traditionally differed from tabling systems. Magic style systems, such as LDL[CGK 90] CORAL [RSS92b] Glue Nail [DMP93] and Aditi [VRK 94] are built upon set at a time engines, and can use set at a time operations like relational joins that may be made efficient for disk resident data, while tabling systems, such as XSB [SSW94] use a tuple at a time strategy that reflects their genesis ....

....problems, but available tabling systems such as XSB [SSW 97] still used a tuple at a time strategy. Even though caching techniques may alleviate the problem for non recursive queries, for recursive queries it is not known in advance how much should be cached. Bottom up systems such as Coral [RSS92b] Aditi [VRK 94] and LDL [CGK 90] can evaluate recursive queries in a set at a time fashion. However, the experimental results of [SW94b, SSW94] have shown that for in memory queries, these systems perform rather poorly compared to XSB. A number of factors contribute to the better ....

[Article contains additional citation context not shown here]

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proceedings of VLDB, pages 238--250, 1992.

....of fixed computation rules, is that if a program is modularly stratified, then it is statically reorderable so that it is modularly stratified for a fixed computation rule. This property has made modular stratification useful for deductive database systems based on magic set evaluation (e.g. [10]) However, whether a program is modularly stratifiable at all is undecidable. Furthermore, programs which can be evaluated using a fixed computation rule may not be modularly stratified, as the program in the following example. Example 1.2 The following program is neither modularly stratified, ....

....of all normal programs. Despite the fact that the results in this paper have been presented using (variants of) SLG resolution, we believe that the underlying ideas are potentially applicable to deductive database systems whose engines have radically different designs from XSB. For example, CORAL [10] evaluates modularly stratified programs using the technique of Ordered Search [9] Described simply, Ordered Search keeps a tree of dependencies and restricts magic evaluation to magic facts that are in an independent SCC. Both Ordered Search and its extension Well Founded Ordered Search as ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL --- Control, Relations, and Logic. In Proceedings of the 18th VLDB Conference, pages 238--249, British Columbia, Canada, 1992.

....which leads to unnecessary recomputation of known derivations. However, it introduces an interesting way of computing the fr(G; C) sets. On the other hand there are several proposals for computing the wellfounded model bottom up. They are usually based on the alternating fixpoint technique, e.g. [15]. The only exception is the Well founded Ordered Search technique [22] Note that while the Well founded Ordered Search restricts the use of the alternating fixpoint to minimum, it can not avoid it completely. Our technique avoids the alternating fixpoint computation altogether by introducing the ....

Ramakrishnan, R., Srivastava, D., Sudarshan, S. CORAL: Control, relations, and logic. Proc. 18th VLDB , 238--249, 1992.

....XSB offers flexible indexing and data storage facilities that make it suitable for a wide range of data oriented problems. It is available through anonymous ftp. 1 Introduction Over the last decade various researchers have explored implementations of logic for data oriented queries [6] 7] [14], 29] Despite its efficient engine Prolog has not always proven useful in these contexts. It is succeptible to infinite loops in programs like transitive closure. As is well known, the ancestor 2 relation will never terminate for Datalog in its left recursive form. In its rightrecursive form it ....

....a top down evaluation in a bottom up framework. However producing an efficient bottom up engine, that moreover can be smoothly integrated with the procedural power of Prolog is a difficult enterprise that we believe has not yet been completely accomplished, despite such noteworthy efforts as CORAL [14], LDL [6] and GLUE NAIL [7] XSB offers an alternative approach to deductive databases. Its use of tuple at a time SLG resolution [5] 3] makes it complete for non floundering programs with finite models, whether they are stratified or not. 1 SLG resolution is currently implemented in the ....

[Article contains additional citation context not shown here]

R. Ramakirshnan, D. Srivastava, and S. Sudarshan. Coral: Control, relations, and logic. In Proceedings of the 18th International Conference on Very Large Data Bases, pages 238--249. VLDB Endowment, 1992.

....for XSB. Within XSB itself, it includes comparisons of SLG derivation to SLD and SLDNF, and gives preliminary indications of the effectiveness of optimizations that are either presently contained in XSB or under consideration for inclusion. It also contains an extended comparison of XSB with CORAL [11]. It concludes with instruction level profiles of XSB which shed light on the behavior observed in the previous sections and also indicate scope for further engine and compiler optimizations. 1 Preliminaries A theory is something nobody believes, except the person who made it. An experiment is ....

....such as EKS IV[ and NAIL[6] are not publicly available, and testing is difficult. In the case of NAIL, tests in [3] on benches involving negation, indicate that the SLG meta interpreter obtains similar times to published times for NAIL on similar platforms. We have not yet obtained a copy of LDL. [11] presents a preliminary comparison of CORAL with LDL, and indicate that CORAL is usually significantly faster than LDL. If so, then the following tests imply that XSB will be significantly faster than LDL as well. 3.1 Comparisons with CORAL A pre release of XSB version 1.3 was benched against ....

R. Ramakirshnan, D. Srivastava, and S. Sudarshan. Coral: Control, relations, and logic. In Proceedings of the 18th International Conference on Very Large Data Bases, pages 238--249. VLDB Endowment, 1992.

....possible queries to a program. As a result, a persistent theme of logic programming has been to investigate how to combine the advantages of bottom up evaluation with the goal orientation of top down techniques. This effort has led to many systems based on magic evaluation and related strategies [4, 13, 19, 20]. The raw speed of top down engines, though is sometimes neglected in the literature. At least for loop free, stratified programs with few redundant subcomputations, top down engines can be substantially faster than bottom up. Thus, rather than adding goal orientation to a bottom up engine, a ....

R. Ramakrishnan, D. Srivastava, and S. Sudarshan. CORAL: Control, relations, and logic. In Proceedings of the Eighteenth International Conference on Very Large Databases, pages 238--249, 1992.

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