S. Finkelstein. "Common expression analysis in database applications." In Proc. SIGMOD, 1982.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....for speeding up subsequent query processing has been studied widely in previous literature. The benefit of this technique is obtained from saving (part of) the subsequent query computations by utilizing the previous cached (intermediate) results. Applications of this technique can be found in [Fin82, LY85, DR92, Sel87, Rou91, Jhi88, HS93, AL80] In [Fin82, LY85, Rou91] cached query results were used in relational database systems to avoid repeated computations. Sel87, Jhi88] addressed the problem This research was sponsored partially by NSF under grant IRI 9057573 and GDR 85 00108 and by ....

....studied widely in previous literature. The benefit of this technique is obtained from saving (part of) the subsequent query computations by utilizing the previous cached (intermediate) results. Applications of this technique can be found in [Fin82, LY85, DR92, Sel87, Rou91, Jhi88, HS93, AL80] In [Fin82, LY85, Rou91] cached query results were used in relational database systems to avoid repeated computations. Sel87, Jhi88] addressed the problem This research was sponsored partially by NSF under grant IRI 9057573 and GDR 85 00108 and by NASA USRA under contract FCPO 550 81. This paper also ....

[Article contains additional citation context not shown here]

S. Finkelstein. Common expression analysis in database applications. In Procs. of ACM-SIGMOD, pages 235--245, 1982.

....optimize evaluation in batch of a set of queries. The developed techniques are geared towards finding and reusing common sub expressions in the set of queries and are heuristics based. The idea of the caching of query results to optimize the processing of subsequent queries was first studied in [11] and [17] The developed techniques are restricted to a subset of extensional, conjunctive queries. No self joins are permitted. The techniques do not, however, find queries that are contained by the original query. In [8] the ADMS system is described, which includes a query caching system ....

....[17] The developed techniques are restricted to a subset of extensional, conjunctive queries. No self joins are permitted. The techniques do not, however, find queries that are contained by the original query. In [8] the ADMS system is described, which includes a query caching system based on [11]. Both [9] and [15] extend the paradigm of query caching to use caches to provide partial answers to the query. They assume, however, that a semantic cache is only useful when some of the query s answers can be obtained from a single cache via project and select operations. Although this framework ....

S. Finkelstein. Common expression analysis in database application. In Proceedings of SIGMOD, pages 235--245, 1982. P. Godfrey & J. Gryz

....management and reuse of (partial) answers is left to the user, who explicitly manage their retention and the construction of access paths to improve response time of recurring and overlapping queries. Few papers have been published on browsing sessions and their optimization. An early paper is [3][6], 1 This work is supported in part by the Pythagoras project (ESPRIT III 7091) which describes a method for reusing answers to previous queries based on identifying common subexpressions. Our approach extends this work by also considering intermediate results produced during query evaluation for ....

....evaluation plan and decide on the answers to be retained. The former task is supported by the query dependency graph. The new query is added to the graph and the BSO selects a subset of the relationships generated by this action for reuse. An algorithm to accomplish this task can be found in [6]. The BSO relies on heuristics to identify reuseable answers, because their future use is unknown. The heuristics take into consideration the query answer and useful intermediate results. The latter requires a large class of subqueries to be considered as well. Several strategies for this are ....

[Article contains additional citation context not shown here]

S. Finkelstein: "Common Expression Analysis in Database Applications", ACM SIGMOD Record, 1982,p235-245.

....limited attention in the database research community. As query optimization was shown to be NP complete [IK84, SM97] it is not surprising that the problem of MQO is also NP complete [SG90] MQO can therefor only be achieved using heuristics [Jar85] or probabilistic techniques. Early works [Fin82, Sel88] show that ad hoc queries can bene t from using materialized results generated by earlier queries, even if only equivalent expressions are considered. The savings can be considerable when compared to single query processing. Shim et al. SSN94] propose improved heuristics to search for the ....

S. J. Finkelstein. Common Expression Analysis in Database Applications. In Proc. of the ACM SIGMOD Int'l. Conf. on Management of Data, pages 235-245, Orlando, FL, USA, June 1982.

....is that our results also include soundness and completeness theorems. The problem of simultaneously optimizing and merging a set of queries has been studied within the context of relational and deductive databases [Grant and Minker 1980; Sellis 1988; Shim et al. 1994; Sellis and Ghosh 1990; Finkelstein 1982; Chakravarthy and Minker 1985] Grant and Minker 1980; Sellis 1988; Sellis and Ghosh 1990; Shim et al. 1994] address the problem of creating a globally optimal access plan for a set of queries, provided that the common expressions among the queries are given as input. Grant and Minker 1980] ....

....our merging algorithms are much faster than the A based algorithm. As the A based algorithm examines a larger search space, it may find low cost plans that our merging algorithms may miss. However, the time it takes to find such good plans is usually not offset by the savings it achieves. [Finkelstein 1982; Chakravarthy and Minker 1985] on the other hand, focus on detecting common expressions among a set of queries in relational and deductive databases. Since the notion of common subexpression varies for different data sources, the common expression identification problem for agents is very ....

Finkelstein, S. 1982. Common expression analysis in database applications. In Proc. of the ACM SIGMOD Conf. on Management of Data (Orlando, Florida, USA, 1982).

....also want information about data value distribution. These flows are not included now. c. An optimizer can maintain transient information about buffers and data caches likely to be available. Reuse of recently retrieved information can increase the efficiency of query processing greatly [Fink 82] 4.M311. Single Relation Operation Executor (L3) a.The Single Relation Operation Executor performs the selection operation on an individual relation, resulting in a record stream. The selection operation uses available indexes implemented by the File Access Module (M101) so that only subsets ....

S.J. Finkelstein: "Common Expression Analysis in Database Applications"; ACMSIGMOD 82, Schkolnick(ed), Orlando FL, Jun.1982, pp.235--245.

....2 E i e i;j 2 E j 2 4. Exploiting interdependencies between queries 9 q 2 3 e 1,2 q 1 e 2,3 q Figure 4: Relationship graph corresponding to Figure 2 The detection of common subqueries is beyond the scope of this paper. Several algorithms have been proposed to detect common subqueries [Fink82, RoHu80]. In the remainder of this paper, we assume that a common subquery can be generated. 4.2 Approach Before presenting our approach, we introduce the notion of a common subquery matrix, abbreviated as csq matrix. A csq matrix for a sequence of n queries has size of (n Gamma 1) by (n Gamma 1) An ....

....implementation of the parts of each step that is not straightforward and for which no algorithms are described in literature. The core of step 1 is to build a csq matrix with regard to the queries of S. We have already noticed that a csq matrix can be generated by using algorithms described in [Fink82, RoHu80]. In Section 4.2.1, we have described how to obtain for each query q its corresponding set Q from the csq matrix. 6. A case study 14 More effort is required for the application of rules 1 and 2 in step 2. Let us describe algorithms to perform these rules. Rule 1 can be applied as follows. A query ....

[Article contains additional citation context not shown here]

Finkelstein, S., Common Expression Analysis in Database Applications, in Proc. of the 1982 ACM Int. Conf. on Management of Data, pp. 235-245.

....First, ECA rules are temporally persistent. That is, they have a longer life span and as a result are likely to be evaluated many times. This suggests that several rules can be optimized simultaneously in a group, possibly using some of the techniques developed for multiple query optimization [Fin82, CM86, RC88, Sel86, Cha91] The effect of multiple query optimization can be further enhanced by materializing intermediate results (e.g. common subexpressions) judiciously. Second, rules used for some applications are likely to have priorities or timing requirements associated with their ....

S. Finkelstein. Common Expression Analysis in Database applications. In Proc. of ACM-SIGMOD, Orlando, Jun. 1982.

....find a globally optimal plan. While there has been work on multi query optimization in the past ( Sel88b, SSN94, SG90, CLS93, PS88] prior work has concentrated primarily on exhaustive algorithms. Other work has concentrated on finding common subexpressions as a post phase to query optimization [Fin82, SV98] but this gives limited scope for cost improvement. We discuss related work in detail in Section 7. The search space for multiquery optimization is doubly exponential in the size of the queries, and exhaustive strategies are therefore impractical; as a result, multi query optimization ....

....so if the query is cheap, yet syntactically complex. For more expensive queries, as well as canned queries that are optimized rarely but executed frequently over large databases, it clearly makes sense to use Greedy. 7 Related Work The multi query optimization problem has been addressed in [Fin82, Sel88b, SSN94, SG90, CLS93, PS88, CR94, ZDNS98, SV98] The work in [Sel88b, SSN94, SG90, CLS93, PS88] describe exhaustive algorithms; they use an abstract representation of a query Q i as a set of alternative plans P i;j , each having a set of tasks t i;j;k , where the tasks may be shared ....

S. Finkelstein. Common expression analysis in database applications. In SIGMOD Intl. Conf. on Management of Data, pages 235--245, Orlando,FL, 1982. 19

....2 confirms that this goal has been met. Generalizations The enumeration algorithm ExOptPlan is robust in that it is completely independent of the algorithm used to generate MapTable. This would make it possible to pick an algorithm for generating equivalent queries using other algorithms [Fin82, YL87, CR94] see discussion in the following section) The optimization algorithm presented in this paper extends to the case where the query and the materialized views are single block Select Project Join queries (i.e. not necessarily conjunctive queries) Most commercial optimize multiblock ....

....choice between the approaches has to be predetermined. Thus, the optimizer can not explore both the options depending on the query and cost estimations. The task of generating equivalent queries based on existing query fragments or semantic knowledge has been studied in several different contexts [Fin82, LY85, YL87, Sel88, CGM90, CS93, CR94] However, all these techniques generate equivalent queries explicitly . In contrast, much of our efficiency in optimization stems from the implicit encoding of the set of equivalent queries in MapTable and a join enumeration algorithm that exploits the ....

S. Finkelstein. Common expression analysis in database applications. In Proceedings of the ACM SIGMOD Conference on the Management of Data, pages 235--245, Orlando, FL, June 1982.

....find a globally optimal plan. While there has been work on multi query optimization in the past ( Sel88b, SSN94, SG90, CLS93, PS88] prior work has concentrated primarily on exhaustive algorithms. Other work has concentrated on finding common subexpressions as a post phase to query optimization [Fin82, SV98] but this gives limited scope for cost improvement. We discuss related work in detail in Section 7. The search space for multiquery optimization is doubly exponential in the size of the queries, and exhaustive strategies are therefore impractical; as a result, multi query optimization ....

....so if the query is also syntactically complex. For more expensive queries, as well as canned queries that are optimized rarely but executed frequently over large databases, it clearly makes sense to use Greedy. 7 Related Work The multi query optimization problem has been addressed in [Fin82, Sel88b, SSN94, SG90, CLS93, PS88, CR94, ZDNS98, SV98] The work in [Sel88b, SSN94, SG90, CLS93, PS88] describe exhaustive algorithms; they use an abstract representation of a query Q i as a set of alternative plans P i;j , each having a set of tasks t i;j;k , where the tasks may be shared ....

S. Finkelstein. Common expression analysis in database applications. In SIGMOD Intl. Conf. on Management of Data, pages 235--245, Orlando,FL, 1982.

....materialized views when they match syntactically a sub expression of the query. In the ADMS optimizer [CR94] subquery expressions corresponding to nodes in the query execution (operator) tree were also cached. A cached result is matched against a new query by using common expression analysis [Fin82]. Semantic matching between a cached view and a (sub)query via mappings, and grouping and aggregation issues, were not addressed. View usability has been studied for conjunctive queries with set semantics and without grouping and aggregation in, e.g. YL87, LMSS95] Levy et al. LMSS95] showed a ....

S. Finkelstein. Common expression analysis in database applications. In Proceedings of the ACM SIGMOD Conference on Management of Data, 1982. 19

....pure relational queries [2] have been specified in accordance to [5] where merging and fragmentation of semantic cache regions is considered. 4. 1 A content based cache description Regarding relational queries the development of a semantic cache descripition is oriented at so called PSJ queries [7,17,19]. Queries which are composed only of projections, selections and joins. In correspondence to those ideas we developed a semantic cache description (predicate representation) and an accordingly matching algorithm (matching of predicates) for the handling of the objects analysis unit. Costs are ....

S. Finkelstein. Common expression analysis in database applications. In Proc. of the ACM SIGMOD Conference, pages 235--245, 1982.

....Given a set fQ 1 ; Q n g of simultaneously posed queries, can we create a query plan that optimizes this set of queries relative to some underlying cost model . The problem of multiple query optimization has been well studied in the relational setting [CLS 95, Cos99, PS88, S88, RH80, Fink82, Jar85, CM85, SSN94] and consists of two parts: i) identifying common subexpressions amongst the queries fQ 1 ; Q n g, and (ii) computing a single, unified global query plan that simultaneously optimizes the total expected cost of the set fQ 1 ; Q n g of queries by using a ....

....optimizes the total expected cost of the set fQ 1 ; Q n g of queries by using a cost model. To date, there is no existing work on the problem of common subexpression analysis in mediated systems, though some work has been done on this hard problem in the relational setting [RH80, Fink82, Jar85, CM85] When describing heterogeneous data sources, what constitutes a common subexpression is unclear. In particular, as mediator platforms like DISCO, GARLIC, HERMES and TSIMMIS, all support integration of new (i.e. never seen before) data sources, the notion of a common ....

[Article contains additional citation context not shown here]

S. Finkelstein,"Common Expression Analysis in Database Applications", In Proc. of ACM SIGMOD'82, Orlando, Florida, 1982

....functions can be generated automatically or can be written by an organization s naming administrator. Performance is further improved in Nomenclator by caching query responses and the descriptions of data distribution. Nomenclator extends relational database work in multiple query optimization [8] to data caching for improved descriptive query performance. New queries that are covered by the results of previous queries are answered from the data cache. Nomenclator introduces meta data caching as a technique for improving descriptive query performance. In meta data caching, information ....

....that cover a query are re used to answer new queries. To cover a query, a cached result must have a selection predicate that is implied by the new query. It must also contain all the projected attributes in the new query. Our techniques for using cached query results are similar to Finkelstein s [8]. We extend these techniques to the meta data caching of referrals that describe the distribution of naming data and the conditions for using catalog functions. The caching of name server resource records (NS RR s) in the Domain Name System (DNS) 15] and prefix tables in Sprite [22] are examples ....

S. Finkelstein, "Common Expression Analysis in Database Applications," ACM SIGMOD International Conference on Management of Data, Orlando, FL, pp. 235-245 (June 1982).

.... string of symbols and detects the common subexpressions within a single query by a bottom up traversal procedure [6] Finkelstein shows how an ad hoc query may be improved by comparing an incoming query with materialized results (intermediate results and final answer) produced from earlier queries [4]. He deals only with equivalent expressions. Jarke discusses the common subexpression isolation in relational algebra, domain relational calculus, and tuple relational calculus [8] Chakravarthy identifies the equivalence and subsumption of two expressions at the logical level, using heuristics ....

S. Finkelstein. "Common Expression Analysis in Database Applications," SIGMOD, pp. 235-245, June 1982.

....which data to retrieve and cache. It might appear at first glance that one should simply consider caching the data for which the user s query asked, or some structurally syntactically determined subset of it after all, it is going to be retrieved anyway. This approach is taken in other systems [3, 11, 2]. It is the only reasonable approach available when the only way to index classify the cached data is by the query it corresponds to. However, it raises two problems mentioned earlier: such data is unlikely to be useful except where the identical query is issued repeatedly, and comparing it to new ....

....that the data returned for this query is not even a subset of obtained for the first one the typical situation caching is envisioned for. 6 Related Work The most closely related work on caching in the database literature is the work by Finkelstein on the topic of common subexpression analysis [3]. Finkelstein presents an approach to reusing previously retrieved data to answer queries or subexpressions of queries. His approach is to cache the results of queries and then look for subexpressions of a new query that match classes of previously cached data and reuse that data where ....

Sheldon Finkelstein. Common expression analysis in database applications. In Schkolnick, editor, Proceedings of ACM SIGMOD, pages 235--245, 1982.

....of a piece of another partial solution, then the former contributes less and can therefore be removed from further consideration if it also has a higher cost. Query signatures, an encoding of the names of all the relations used by the query, can be used to perform these comparisons efficiently [8]. It is interesting to see how the new algorithm behaves when it is given a set of gmaps that represents a traditional relational physical schema. Assume for example that one gmap is a file containing the extent of the Faculty relation with all associated attributes, defgmap facultyrelation as ....

....the work on materialized views [3, 28] such efforts include research whose goal was not physical data independence but simply processing efficiency. Examples include research on reusing common subexpressions within a query [9] or between multiple queries [22] reusing results of previous queries [8], and using integrity constraints for semantic query optimization [6] Kemper and Moerkotte [13] opt for a unified approach of translation and opPage timization for the ASRs by extending a rule based optimizer to include appropriate rewriting rules. Our approach of enhancing a conventional ....

S. Finkelstein. Common expression analysis in database applications. In Proc. of the ACM SIGMOD Conf., pages 364--374, 1982.

No context found.

S. Finkelstein. "Common expression analysis in database applications." In Proc. SIGMOD, 1982.

No context found.

S. Finkelstein. Common expression analysis in database applications. In Proc. ACM SIGMOD, 1982.

No context found.

S. Finkelstein. "Common expression analysis in database applications." In Proc. SIGMOD, 1982.

No context found.

Finkelstein, S. (1982) Common expression analysis in database applications. In Schkolnick, M. (ed.), ACM SIGMOD Int. Conf. on Management of Data, Orlando, FL, June 2--4, pp. 235--245. ACM Press, New York.

No context found.

Finkelstein, S., "Common expression analysis in database applications," Proc. 1982.

No context found.

FS82 Finkelstein, S., Common Expression Analysis in Database Applications. Proc. 1982 SIGMOD Conf. on Management of Data, ACM, New York, N.Y., (1982), 235-245.

No context found.

Finkelstein, S., Common Expression Analysis in Database Applications, in Proc. of the 1982 ACM Int. Conf. on Management of Data, ACM Press, pp. 235-245, 1982.

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