A. Gupta, I. S. Mumick, and K. A. Ross. Adapting materialized views after redefinitions. In Proc. ACMSIGMOD, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....namely, View Maintenance (VM) View Synchronization (VS) and View Adaptation (VA) VM [19, 1, 16] mainrains the DW view extent under source data updates. In contrast to VM, VS [8, 12] aims at rewriting the DW view definition when the source schema has been changed. Thereafter, View Adaptation (VA) [13, 7] incrementally adapts the view extent to again match the newly changed view definition. If there is no concurrency among source updates, namely, the current DW maintenance completes before the next source update occurs, then VM incorporates each source data update (DU) while VS and VA together ....

.... Work Maintaining DW materialized views under source updates is an important is sue [19, 6] Besides incremental view maintenance under source data updates [6] there is some work on view rewriting upon source schema changes [8] and on adapting the view extent after the view has been redefined [7, 13]. In a loosely coupled environment, anomaly problems can arise due to autonomous source updates. 19] introduces a compensation based algorithm to handle concurrent data updates restricted to one single source. SWEEP [1] applies local compensation over distributed sources. 16] proposes to ....

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views after Redefinition. In Proceedings of SIGMOD, pages 211-222, 1995.

....for the nondistributive aggregate functions. Other related work includes rendering the optimizer aware of the incremental view maintenance process [Vis98, MRSR01] dealing with the case where the views and the base data are decoupled [ZGMHW95] and maintaining views under structural changes [GMRR01] 7 Conclusions Incremental view maintenance is an extremely important aspect of the modern database management systems. It enables the fast execution of complex queries without sacrificing the freshness of the data. However, the maintenance of views defined with non distributive aggregate ....

Ashish Gupta, Inderpal S. Mumick, Jun Rao, and Kenneth A. Ross. Adapting materialized views after redefinitions: Techniques and a performance study. Information Systems, 26(5):323--362, July 2001.

....new information can be added but deletions and modifications are either forbidden or subject to several restrictions. In the relational database framework, much work on update propagation from view relations to base relations [BS81, DB82, Kel85] and maintenance of materialized views [GKM92, GMR95, CW91] has been done to date. Nevertheless, commercial relational databases systems apply the most simplistic and pragmatic approaches. Updates can be performed only on views defined by simple query expressions, e.g. queries involving only one base relation. Such queries involve selection and ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting Materialized Views after Redefinitions. In Proc. ACM SIGMOD Symp. on the Management of Data, pages 211--222, 1995.

....second research area called view synchronization (VS) NLR98, RLN97, LNR97] is concerned with evolving the view de nition itself whenever there is a schema change (SC) of one of the ISs that results in a view de nition to become unde ned. The third research area, referred as view adaptation (VA) GMR97, MD96, NR99] is concerned with adapting the view extent incrementally after the view de nition has been modi ed either directly by the user or indirectly by a view synchronization module. Among these three areas, View maintenance (VM) is the only one that has given attention to the problem of ....

A. Gupta, I. S. Mumick, and J. Rao. Adapting Materialized Views after Redefinitions: Techniques and a Performance Study. Technical Report CUCS-010-97, Columbia University, 1997.

....as A(t) We can recursively compute the two sets from queries on the metadata repository of process definitions. Queries for the successor and after relationships can be defined in a similar way. Suppose that the final SQL expression of a type t, say e, changes into e . In the spirit of [24], we can use the following rules for schema evolution in a data warehouse environment (we consider that the changes abide by the SQL syntax and the new expression is valid) If the select clause of e has an extra attribute from e, then propagate the extra attribute to the base relations: ....

....implicit impacts, we do not provide a fully automated algorithmic solution to the problem, but rather, we sketch a methodological set of steps, in the form of suggested actions to perform this kind of evolution. Similar algorithms for the evolution of views in data warehouses can be found in [24, 4]. A tool could easily visualize this evolution plan and allow the user to react to it. 6. Related Work In this section we discuss the state of art and practice for research efforts, commercial tools and standards in the fields of process and workflow modeling, with particular focus to data ....

A. Gupta, I. Mumick, K. Ross. Adapting Materialized Views after Redefinitions. In Proc. ACM SIGMOD Intl. Conf. on the Management of Data, pp. 211-222, San Jose, CA (1995).

....materialized and the view is redefined by changing its definition slightly. If the second view is also going to be materialized, it might be possible to use the old value of the view and adapt it to conform to the view s new definition appropriately. This problem has been studied by Gupta et al. GMR95] 4. Answering Queries Using Views If views are materialized, the query processor might be able to use this set of materialized views, in order to answer other queries [LMSS95, FRV96] In general, this problem is difficult, but a solution to it might be very useful, especially in applications ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting Materialized Views after Redefinitions. In Proceeding of ACM-SIGMOD Conference on Management of Data, pages 211--222, 1995. BIBLIOGRAPHY 120

.... methods for view maintenance have been proposed [15, 53, 141] View redefinition refers to a change in a view definition; the usual way to update a view upon redefinition is to compute the new view from scratch; a more efficient alternative may be to exploit existing views through view adaptation [54]. 38 4.3 The Dye View Definition Language Dye is a database view definition language which we propose for the purpose of defining and maintaining active views of hygraphs. The Dye language is an object oriented algebra that employs schema preserving relational algebra operations 1 with ....

....the display (FILTER) The naive way to evaluate the dynamic selection query is simply to compute the result from scratch. A more efficient approach is to view this problem as an instance of view redefinition, which permits us to exploit the result of the previous query through view adaptation [54]. In particular, we can keep track of the set of visible objects and the set of masked objects; two scenarios are then possible: ffl If a DQ component is manipulated to narrow the range of the selection, we evaluate the selection query against the set of currently visible objects; objects that do ....

[Article contains additional citation context not shown here]

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting Materialized Views after Redefinitions. ACM SIGMOD Record, 24(2):211--222, June 1995.

....paper, we now address the new problem that after the view synchronization process modifies the view definition, the view extents if materialized must be brought up to date as well. This problem is similar to the explicit view redefinition problem that has recently been studied in the literature [7, 4]. That is, the view definition changes triggered by the IS changes could be mapped into a sequence of simple primitive changes of different types assumed requested by a user to be executed explicitly on the view definition [7, 4] Once this mapping is established, one can apply the ....

.... redefinition problem that has recently been studied in the literature [7, 4] That is, the view definition changes triggered by the IS changes could be mapped into a sequence of simple primitive changes of different types assumed requested by a user to be executed explicitly on the view definition [7, 4]. Once this mapping is established, one can apply the maintenance after redefinition strategies proposed in [7, 4] for each such simple change. However, as we will demonstrate, this strategy is very inefficient, as in the case of view synchronization the changes done to a view definition are ....

[Article contains additional citation context not shown here]

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views after Redefinition. In Proc. of ACM SIGMOD International Conf. on Management of Data, pages 211--222, 1995.

....problem arises due to the changing user requirements. As the requirements of the user changes, the view definitions at the data warehouse change dynamically. There is a need to develop view adaptation techniques that do not need the entire recomputation of the materialized view at every change. In [26, 39, 38], the authors have presented the view adaptation techniques for a data warehouse that recompute only parts of the view (if any) which cannot be derived from the existing materialized views. These techniques are applicable for SPJ queries. However, there is a need to devise adaptation algorithms ....

A. Gupta, I.S. Mumick and K.A. Ross. Adapting materialized views after redefinitions. In Proc. ACM SIGMOD International Conference on Management of Data, San Jose, USA, 1995.

....occurs at the IS space. The second area called view synchronization (VS) 9, 10] is concerned with evolving the view definition itself whenever there is a schema change (SC) of one of the ISs that results in a view definition to become undefined. The third area, referred to as view adaptation (VA) [2, 6, 7], adapts the view extent incrementally after the view definition has been changed either directly by the user or indirectly by a view synchronization module. Materialized view maintenance (VM) is the only area among those three that thus far has given attention to the problem of concurrency of ....

....from IS[i] at sequence number n. Q(n) Query used to handle update X(n) i] Q(n) i] Sub query of Q(n) sent to IS[i] QR(n) i] Query result of Q(n) i] QR(n) Query result of Q(n) after reassembly of all QR(n) i] for all i. Fig. 2. Notations and Their Meanings. QR(1) SC(1) 1] QR Q(1) DU(2)[2] IS[2] DW Q(1) 1] SC DU IS[1] Q Q(1) 2] Q QR(1) 1] QR QR(1) 2] Fig. 3. Time Line for a Maintenance Concurrent Data Update. Figure 3 illustrates the concept of a maintenance concurrent update defined in Definition 1 with a time line. Messages only get time stamps assigned upon their ....

[Article contains additional citation context not shown here]

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views after Redefinition. In Proceedings of ACM SIGMOD International Conference on Management of Data, pages 211--222, 1995.

.... by set differences [BM95b] View indexing by means of pointer structures has been described in [Baek95] BM95a] BM95b] BM95c] JMR91] QW91] Rou82a] Rou82b] Rou91] RK91] Val87] The use and refreshment of materialized views has been de2 scribed in [BM90] CKPS95] GL95] GMR95] HK95] Han87] LMSS95] SJGP90] SR87] ZGMHW95] Incremental techniques has been described for active databases [CSL 90] CW91] HCKW90] Han92] RCBB89] WDSY91] for deductive databases [GMS93] and for temporal databases [BM95c] JMR91] McK88] However, we are not aware of ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting Materialized Views after Redefinitions. In Proceedings of the 1995 ACM SIGMOD International Conference on Management of Data, San Jose, California, USA, pages 211--222, 1995.

....auxiliary SQL expressions producing, thus the set e= e 1 ,e 2 , e . Obviously some of the expressions belonging to e belong also to B(t) Thus, we extend B(t) as B(t)e (with set semantics) Suppose, then, that the final SQL expression of a type t, say e, changes into e . Following the spirit of [GuMR95], we can use the following rules for schema evolution in a DW environment (we consider that the changes abide by the SQL syntax and the new expression is valid) If the select clause of e has an extra attribute from e, then propagate the extra attribute down the line to the base relations: ....

....for the successor objects. We do not claim that we provide a concrete algorithmic solution to the problem. Rather, we sketch a methodological set of steps, in the form of suggested actions to perform this kind of evolution. Similar algorithms for the evolution of views in DW s can be found in [GuMR95, Bell98]. A tool could easily visualize this evolution plan and allow the userto react to it. 5 Related Work In the field of workflow management, WMC98] is a standard proposed by the Workflow Management Coalition (WfMC) The standard includes a metamodel for the description of a workflow process ....

A. Gupta, I. Mumick, K. Ross. Adapting Materialized Views after Redefinitions. In ACM SIGMOD Conference, 1995.

.... database systems, such as mediators (see [21, 18] and, more in general, the development of tools for supporting integrated access to cooperative information systems [9, 13] 2) query optimization [7] and view maintenance [10] 3) structuring and maintenance of warehouses and constraints [15]. As an example we describe how extracted patterns can be used for materialized view structuring; the problem is the following: given a database, we want to decide which views to materialize for the purpose of optimized query answering. Now, PDL formulae naturally correspond to queries, and then ....

A. Gupta, I.S. Mumick, K.A. Ross, Adapting materialized views after redefinitions. Proc. ACM SIGMOD Conf., 1995

....the relations may be distributed or replicated, and locating as well as accessing them may be expensive. Locally cached materialized views of the data, such as the results of previous queries, may improve the performance of such applications. Data warehousing [GJM96, ZGMHW95] view adaptation [GMR95] and physical data independence [TSI94] involve the usage of materialized views (conceptual relations) to integrate data from multiple sources or to hide schema differences or physical storage details. The underlying relations may be expensive to access or may not be available, requiring queries ....

....views to answer aggregation queries; their approach uses transformation rules on the query tree representations, does not provide any formal guarantees of completeness, and does not deal with many of the cases covered by our approach (such as Example 1. 1) A restricted problem is studied in [GMR95], who assume that a materialized view may be redefined, and investigate how to adapt the materialization of the view to reflect the redefinition. A comparison with [CKPS95, GHQ95, GMR95] and other related work in the literature is presented in Section 6. 1.1 Motivating Example We present an ....

[Article contains additional citation context not shown here]

Ashish Gupta, Inderpal S. Mumick, and Kenneth Ross. Adapting materialized views after redefinitions. In Proceedings of the ACM SIGMOD Conference on Management of Data, San Jose, CA, May 1995.

....of evolution processes. As an example, we will discuss the evolution of views in data warehouses. The evolution of data warehouse views has been studied recently in the research fields of schema evolution [RLN97, Bell98, Blas99] and maintenance of data warehouse views under view redefinition [GMR95]. In this section, we do not provide a new technique for schema evolution or view maintenance of data Materialize View StoreView MetaData StoreView Extent Evaluate View AddDW Relation composed Of next next next affects DW Completeness DataStore Availability DataStore Minimality ....

....The framework supports among other features the automatic adaptation of instances, change notification for applications, and forward compatibility of schemata. The second perspective which addresses the problem of evolution of data warehouse views, is maintenance of the extent of a view. In [GMR95], the problem of incremental view maintenance under view redefinition is studied. An overview and a taxonomy of view maintenance problems is given in [GM95] HMV99] studies the problem of maintaining multi dimensional data cubes under dimension updates. They define a basic set of operators why ....

A. Gupta, I.S. Mumick, K.A. Ross. Adapting Materialized Views after Redefinitions. In Proc. ACM SIGMOD International Conference on Management of Data, pp. 211--222, 1995.

....research area called view synchronization (VS) NLR98, RLN97, LNR97] is concerned with evolving the view definition itself whenever there is a schema change (SC) of one of the ISs that results in a view definition to become undefined. The third research area, referred to as view adaptation (VA) GMR97, MD96, NR99] is concerned with adapting the view extent incrementally after the view definition has been changed either directly by the user or indirectly by a view synchronization module. Materialized view maintenance (VM) is the only area among those three that thus far has given attention to ....

....query, because the query result cannot be computed by IS2. The data warehouse can no longer be maintained correctly because the incremental view maintenance process is based on obtaining the results of the maintenance queries. A similar broken query problem also holds for view adaptation queries [GMR97, NR99] as they are also send down to the IS space (or for that matter for any query send down to the IS space) and thus face the issue of the IS schema changing unexpectedly. 1.3 Our Solution Approach The DyDa Framework In this paper, we present a general approach called the Dynamic Data ....

[Article contains additional citation context not shown here]

A. Gupta, I. S. Mumick, and J. Rao. Adapting Materialized Views after Redefinitions: Techniques and a Performance Study. Technical Report CUCS-010-97, Columbia University, 1997. 32

....such problem arises due to the changing user requirements. As the requirements of the user changes, the view definitions at the data warehouse change dynamically. We need to develop view adaptation techniques that do not need the entire recomputation of the materialized view at every change. In [24, 37, 36], the authors have presented the view adaptation techniques for a data warehouse that recompute only parts of the view (if any) which cannot be derived from the existing materialized views. These techniques are applicable for SPJ queries. However, there is a need to devise adaptation algorithms ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting materialized views after redefinitions. In Proc. ACM SIGMOD International Conference on Management of Data, San Jose, USA, 1995.

.... queries using views has recently received considerable attention because of its applications in mediator systems (e.g. Information Manifold [LRO96] TSIMMIS [CGMH 94, PGGMU95] SIMS [AKS96] HERMES [ACPS96] Razor [FW97] Infomaster [DG97] Mobile Computing [BI94, HSW94] view adaptation [GMR95] maintaining physical data independence [TSI96] and speeding up query processing [YL87, CKPS95] The problem arises naturally in mediator systems that provide access to multiple heterogeneous information sources [Ull97] In such systems, The work was done while the first author was at AT T ....

Ashish Gupta, Inderpal Singh Mumick, and Kenneth A. Ross. Adapting materialized views after redefinitions. In Proceedings of SIGMOD-95, 1995.

....A naive approach would take the new query and would submit it to the system. A more sophisticated approach can take advantage of an already processed query by incrementally modifying its result. Incremental approaches have been successfully used in database systems for view maintenance [CW91, GMR95, GL95] In most situations, an incremental approach is substantially more efficient than a naive one. Efficiency improvement is more visible when several consecutive changes to the query are required, i.e. during iterative browsing, and when a large case base is used. Small updates to the query ....

....several consecutive changes to the query are required, i.e. during iterative browsing, and when a large case base is used. Small updates to the query generally produce only small changes to the query result. Thus, only local changes to the query are required if an incremental approach is used [GMR95] The basic idea of incremental computation in TA3 is to store query results and reuse them when similar queries are computed, similarly as in [BM95] If the number of attributes per case is significantly smaller than the number of cases then the incremental context restriction relaxation ....

A. Gupta, I. S. Mumick, and K. A. Ross. Adapting materialized views after redefinitions. In Proc. of the ACM SIGMOD International Conference on Management of Data, pages 211--222, San Jose, CA, 1995.

....correspondences between the types of nodes that were considered in VDP merging. If there is a correspondence between a node n in the old VDP and a node n 0 in the new VDP, the data of class(n 0 ) can be derived (perhaps partially) from the data of class(n) using techniques developed in [GMR95] that adapt a view in response to changes to the view definition. Those techniques primarily use existing data in the materialized view with minimal access to the source classes. 5 A Taxonomy of the Solution Space for Data Integration In its current form, the Squirrel system can be used to ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting materialized views after redefinition. In Proc. ACM SIGMOD Symp. on the Management of Data, pages 211--222, San Jose, CA, May 1995.

.... changes [LNR97a, LNR97b] View synchronization is in contrast to the large body of work on incremental view maintenance that addresses changes at the data but not at the schema level [ZGMHW95] and to recent work on view redefinition that again focuses on how to efficiently update the view data [GMR95] Previous work on query optimization [vdBK94, AAS97, BLT86] and rewriting view queries [LMS95] was restricted to always requiring exact equivalence between the original and replacement queries. This condition is not likely to be always achievable in a dynamic environment such as the WWW. We have ....

....being implemented in the EVE system. This new implementation of EVE will allow us to efficiently handle schema evolution in a distributed data warehousing environment, which is a significant improvement over current technology handling only data updates at the underlying information sources [GMR95, vdBK94, AAS97, BLT86] Acknowledgments. The authors would like to thank students at the Database Systems Research Group at WPI for their interactions and feedback on this research. In particular, we are grateful to Xin Zhang, Yong Li, and Amber Van Wyk for implementing several of the major ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting Materialized Views after Redefinition. In Proceedings of ACM SIGMOD International Conference on Management of Data, pages 211--222, 1995.

....such problem arises due to the changing user requirements. As the requirements of the user changes, the view definitions at the data warehouse change dynamically. We need to develop view adaptation techniques that do not need the entire recomputation of the materialized view at every change. In [GMR95, MD96, Moh97], the authors have presented the view adaptation techniques for a data warehouse that recompute only parts of the view (if any) which cannot be derived from the existing materialized views. These techniques are applicable for SPJ queries. However, there is a need to devise adaptation algorithms ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting materialized views after redefinitions. In Proc. ACM SIGMOD International Conference on Management of Data, San Jose, USA, 1995.

....our industrial sponsors, in particular, IBM for the IBM partnership award and for the IBM corporate fellowship for one of her graduate students. how the view synchronization problem is different from maintenance after data updates [GM95] or after explicit user specified redefinition of the view [GMR95, MD96] Our Evolvable View Environment EVE approach [RLN97, LNR97a, NLR98] focuses on the problem of data warehouse maintenance when base relations exhibit schema level changes. In our previous work [LNR97b, NLR98, NR98a] we proposed different strategies for rewriting view definitions triggered ....

....views are materialized at the data warehouse site, then after the view synchronization process modifies the view definition the view extents must be brought up to date as well. This problem is similar to the explicit view redefinition problem that has recently been studied in the literature [MD96, GMR95] That is, the view definition changes triggered by the IS changes could be mapped into a sequence of primitive changes assumed to be explicitly requested by a user for the view definition [MD96, GMR95] Once this mapping is established, one can apply the maintenance after redefinition strategies ....

[Article contains additional citation context not shown here]

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting Materialized Views after Redefinition. In Proceedings of ACM SIGMOD International Conference on Management of Data, pages 211--222, 1995.

....modify their data contents, but also their schemas, their interfaces, as well as their query capabilities. Practically all research in view maintenance thus far has focused on the propagation of data updates from ISs to the warehouse [1, 23, 22, 21, 2, 17, 3, 10] Two exceptions are Gupta et al. [5] and Mohania et al. 9] who both propose algorithms for how to keep a materialized view extent up to date when the view definition itself is explicitly changed by the user (for example, when a user drops one attribute from the SELECT clause of a view definition. To the best of our knowledge, the ....

....the concurrency of data updates at ISs, while this paper is the first to address the concurrency problem between data updates and schema changes in such environments. 1.2. Example of Problem: Data Warehouse Maintenance over Evolving ISs We now illustrate with an example how current technology [1, 5, 22] would fail to handle this problem of concurrent data updates and schema changes. Assume we have two information sources IS1 and IS2 with relations R and S, respectively. Figure 1 (initial state) shows the initial extent of R and S, with the view V of the data warehouse defined by the following ....

[Article contains additional citation context not shown here]

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views after Redefinition. In Proceedings of ACM SIGMOD International Conference on Management of Data, pages 211--222, 1995.

.... derive count on a view indicates the number of derivations of each view tuple. By associating this extra information to base relations and views, which takes very little extra space and can be maintained ef ficiently, the total communication cost can be reduced significantly. Gupta et al. [9] present view adaptation methods for changes to SPJ queries in centralised databases. These methods handle only a limited number of changes, and do not consider communication cost, which becomes an important issue when data warehouse systems are considered. The view adaptation problem is related ....

....to view adaptation. This is because view maintenance involves only propagation of updates on base relations to views, where as view adaptation involves processing of changes to view definitions, and computation of the resulting changes to view instances. In this paper, we either extend the work of [9] for data warehouses or propose more efficient methods. The outline of the rest of the paper is as follows. Firstly we introduce some notations, definitions, and our motivating example in section 2. We describe the view adaptation algorithm for changing the SELECT clause in section 3. In 4, we ....

[Article contains additional citation context not shown here]

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting materialized views after redefinitions. In Proc. ACM SIGMOD International Conference on Management of Data, San Jose, USA, 1995.

....will contain all the information that is needed to adapt the view. Maintenance of intermediate results (called auxiliary relations) has also been recommended in data warehouses for making views self maintainable [RSS96, QGMW96] 4 Related Work Recently, view adaptation methods have been proposed [GMR95, MD96b, MD96a]. These methods are restricted to changes in views defined by simple Select ProjectJoin (SPJ) queries. While SPJ queries are simpler to work with than general queries, many realistic view definitions are complex, and also include EXCEPT and UNION. Complex view definitions are particularly ....

....changes in views defined by simple Select ProjectJoin (SPJ) queries. While SPJ queries are simpler to work with than general queries, many realistic view definitions are complex, and also include EXCEPT and UNION. Complex view definitions are particularly prevalent in data warehouses. Gupta et al. [GMR95] present view adaptation methods for changes to SPJ queries in centralised databases. These methods handle only a limited number of changes, and do not consider communication costs, which becomes an important issue when data warehouse systems are considered. Mohania and Dong [MD96b, MD96a] have ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting materialized views after redefinitions. In Proc. ACM SIGMOD International Conference on Management of Data, San Jose, USA, 1995.

....Section 3.3. One type of control change we do not cover in Section 3.3 are changes to V j [S] which occur when V j tuples are expunged. V j [S] can be changed only by adding selection conditions or adding semi joins with other views. The expunged tuples can be found using techniques outlined in [GMR95] Once found, they are treated as deleted V j tuples and propagated to the higher level views. Before we discuss the algorithms, we consider the rule V 8 (X) Gamma V 7 (X; Y ) X 7 and introduce some notation. The variables that appear in the head predicate (i.e. X) are called ....

....results (similar to our Q[A] of a query Q) Another problem that our framework tackled is coping with control changes. In our framework, what is available for view maintenance and querying changes with the controls. We have developed algorithms that cope with control changes. The algorithms in [GMR95] can also be used to find 5 e V j . Apart from [GMR95] and this paper, there has been no work on compensating for control changes. 6 Conclusions We have presented a framework and design for system managed removal of warehouse data. Within it, the warehouse administrator (WHA) provides ....

[Article contains additional citation context not shown here]

A. Gupta, I. Mumick, and K. Ross. Adapting materialized views after redefinitions. In Proceedings of 1995 ACM SIGMOD, pages 211--222, 1995.

....to answer queries. Aside from its potential of improving performance of query evaluation [LY85, YL87, KB94, CKPS95] the ability to use views is important in other applications. For example, in applications such as Global Information Systems [LSK95] Mobile Computing [BI94, HSW94] view adaptation [GMR95], maintaining physical 2 Chapter 1 data independence [TSI94] the relations mentioned in the query may either not actually be physically stored (e.g. they may be only conceptual relations) or be impossible to consult (e.g. they are stored in a remote server that is temporarily unavailable to a ....

Ashish Gupta, Inderpal Singh Mumick, and Kenneth A. Ross. Adapting Materialized Views after Redefinitions. In Proceedings of SIGMOD-95, 1995.

....clear benefit in OLAP and data mining environments. Other application domains include, e.g. scientific databases [HQGW93] and network monitoring systems. Our algorithms and results for lineage tracing also can be applied to the problems of view update [DB78] materialized view schema evolution [GMR95] and data cleansing. See [CWW97] for further discussion on these applications. To compute the lineage of a view data item, we need the view definition and the original source data, as well as possibly auxiliary information representing certain intermediate results in the view definition. In a ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting materialized views after redefinitions. In Proc. of the ACM SIGMOD International Conference on Management of Data, pages 211--222, San Jose, California, May 1995.

....changes of underlying ISs are also permissible changes in the information space. Thus, in the SDCC (Schema change and Data update Concurrency Control) manager [ZR99] we handle concurrency between data and schema changes. SDCC integrates view synchronization [RLN97] view adaptation (Gupta, GMR97] and concurrent view maintenance algorithms for data updates (SWEEP, AAS97] Thus, it allows for data updates and schema updates to occur concurrently and for the data warehouse to adapt itself incrementally to both types of changes. 1.6 View Optimization View optimization addresses the ....

A. Gupta, I. S. Mumick, and J. Rao. Adapting Materialized Views after Redefinitions: Techniques and a Performance Study. Technical Report CUCS-010-97, Columbia University, 1997.

.... methods for view maintenance have been proposed [10, 41, 116] View redefinition refers to a change in a view definition; the usual way to update a view upon redefinition is to compute the new view from scratch; a more efficient alternative may be to exploit existing views through view adaptation [42]. 4.3 Visualization System Architectures In this section we describe how the main components of DQM systems are interrelated through an evolving sequence of high level event action architectures depicted as directed acyclic graphs. In particular, this allows us to situate the view definition ....

....from the display (FILTER) The naive way to evaluate the dynamic selection query is simply to compute the result from scratch. A more efficient approach is to view this problem as an instance of view redefinition, which permits us to exploit the result of the previous query through view adaptation [42]. In particular, we can keep track of the set of displayed objects and the set of masked objects; two scenarios are then possible: ffl If a DQ component is manipulated to narrow the range of the selection, we evaluate the selection query against the set of currently displayed objects; objects that ....

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting Materialized Views after Redefinitions. ACM SIGMOD Record, 24(2):211--222, June 1995.

....upon types of changes that either a view or an information source could undergo. This taxonomy represents a suitable classification based on which we can define as well as differentiate between different view adaption problems, such as materialized view maintenance [GM95, Wid95] view redefinition [GMR95, MD96], 1 These may include information such as their schemas, their query interfaces, as well as other services offered by the information sources. E.A. Rundensteiner, A.J. Lee, A. Nica 13 1 etc. We also use the taxonomy to identify a new view adaptation problem, called view synchronization, that ....

....by (1) the view has no explicit changes apriori, 2) the IS changed its data extent, 3) the view user wants the view extent up to date, and (4) the IS did not preserve its old data. b) The problem of materialized view maintenance after view redefinition has been recently studied by Gupta et al. [GMR95] and by Mohania et al. MD96] Its coordinates are marked by triangles in Figure 2. The main difference to above is that the process is (1) triggered by the view s explicit change of its definition, and (2) no changes at the IS side occurred. c) A new problem materialized view maintenance after ....

[Article contains additional citation context not shown here]

A. Gupta, I.S. Mumick, and K.A. Ross. Adapting Materialized Views after Redefinition. In Proceedings ACM SIGMOD International Conference on Management of Data, pages 211--222, 1995.

....be answered much more quickly by leveraging former results than by evaluating the queries from scratch, especially in a common case where users are exploring data by making small dynamic modifications to pose a sequence of queries. This paper significantly extends a previous article from SIGMOD [GMR95], to which there are already numerous citations (see the DBLP URL, http: www.informatik.uni trier.de ley db conf sigmod sigmod95 16.html) The second paper, by Dimitri Theodoratos, Detecting Redundant Materialized Views in Data Warehouse Evolution, addresses the problem of determining ....

Ashish Gupta, Inderpal Singh Mumick, and Kenneth.A. Ross, "Adapting materialized views after redefinitions," SIGMOD Conference, 1995, 211-222.

....application where an archaeologist tries to discover rules about data by formulating queries, looking at the results (which can be cached) of the query, and then changing the query iteratively as the archaeologist s understanding improves. # A preliminary version of this article appeared as [GMR95] Part of the research was done while the author was working at AT T Laboratories. # The work of Jun Rao was performed while at Columbia University. Research supported by a grant from the AT T Foundation, by a David and Lucile Packard Foundation Fellowship in Science and Engineering, by a ....

Ashish Gupta, Inderpal Singh Mumick, and Kenneth A. Ross. Adapting materialized views after redefinitions. In SIGMOD, pages 211-222, 1995.

....can speed up the process of answering interactive queries. An important practical question is choosing which views to materialize. In [8] we address this problem, demonstrating that it sometimes pays to materialize additional views just to support the maintenance of a given materialized view. In [9] we look at the problem of view adaptation, namely how can one incrementally modify a material ized view after a change in the view definition. When one queries multiple materialized views and base tables one would like to see a single consistent database state. A more general notion of ....

A. Gupta, I. S. Mumick, and K. A. Ross. Adapting materialized views after redefinitions. In Proceedings of the ACM-SIGMOD International Conference on Management of Data, pages 211--222, 1995.

....techniques. Our evaluation indicates that adaptation is more efficient than rematerialization in most cases. Certain adaptation techniques can be up to 1,000 times better. We also point out the physical layouts that can benefit adaptation. A preliminary version of this paper appeared as [GMR95] y Research supported by NSF grants IRI 91 16646 and IRI 92 23405. z Research supported by a grant from the AT T Foundation, by a David and Lucile Packard Foundation Fellowship in Science and Engineering, by a Sloan Foundation Fellowship, by NSF grants IRI 9209029, CDA 90 24735, and by ....

Ashish Gupta, Inderpal Singh Mumick, and Kenneth A. Ross. Adapting materialized views after redefinitions. In SIGMOD, pages 211-222, 1995.

....are two efforts at AT T Bell Laboratories, one investigating how to implement materialized views inside an object oriented database, while the second tries to implement materialized views on top of an existing relational database. A project at Columbia University is implementing view adaptation [GMR95] on top of Sybase. The WHIPS project at Stanford is attempting to integrate data from multiple sources into a data warehouse using materialized views [HGW 95] Acknowledgements I thank Dallan Quass and Timothy Griffin for comments on a draft of this paper, and Ashish Gupta for exploring ....

Ashish Gupta, Inderpal Singh Mumick, and Kenneth A. Ross. Adapting materialized views after redefinitions. In [Sig95].

....is fairly simply expressed as a SELECT DISTINCT over the old view to obtain the new view. Deleting a DISTINCT qualifier is more difficult, since it is not clear how many duplicates of each tuple should be in the new view. Techniques to do so are discussed in the full version of this paper [GMR95] An alternative is to augment the view so as to always keep a count of the number of derivations for each tuple in the view. In this case, changes to the DISTINCT Qualifier can be handled easily by either presenting the count to the user, or by hiding the count. 3.2 Changes in the WHERE Clause ....

....We assume that the initial view definition is as stated at the beginning of Section 3. For each possible redefinition, we give the possible adaptations along with the assumptions needed for the adaptation to work. The assumptions are listed separately in Table 1. In the full version of this paper [GMR95] we also discuss adaptation of SELECT FROM WHERE queries that originally use the DISTINCT qualifier. Table 2 can be used in three ways. Firstly, the query optimizer would use this table to find the adaptation technique (and compute its cost estimate) given the properties of the current schema ....

[Article contains additional citation context not shown here]

Ashish Gupta, Inderpal Singh Mumick, and Kenneth A. Ross. Adapting materialized views after redefinitions. Columbia University Technical Report number CUCS-010-95, March 1995.

....since they do not explore some parts of the full search space. Finally, we discuss possible extensions to our techniques in Section 6. 1. 2 Related Work View maintenance (and the closely related problem of integrity constraint checking) has been studied extensively in the literature (e.g. [2, 4, 8, 11, 12, 18, 22]) for various view definition languages, e.g. Select ProjectJoin (or SPJ) views, views with multiset semantics, views with grouping aggregation, and recursive views; for various types of updates, e.g. insertions, deletions, modifications, to the database relations; and for modifications to the ....

A. Gupta, I. S. Mumick, and K. Ross. Adapting materialized views after redefinitions. In Proceedings of the ACM SIGMOD Conference on Management of Data, San Jose, CA, May 1995.

No context found.

A. Gupta, I. S. Mumick, and K. A. Ross. Adapting materialized views after redefinitions. In Proc. ACMSIGMOD, 1995.

No context found.

A. Gupta, I.S. Mumick, K.A. Ross, Adapting Materialized Views after Redefinitions, SIGMOD Conf., San Jose, 1995.

No context found.

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views After Redefinition. In SIGMOD, pages 211--222, 1995.

No context found.

Ashish Gupta, Inderpal Singh Mumick, and Kenneth A. Ross. Adapting materialized views after redefinitions. In Michael J. Carey and DonovanA. Schneider, editors, Proceedings of the 1995 ACMSIGMOD International Conference on Management of Data, San Jose, California, May 22--25, 1995, pages 211--222. ACM Press, 1995.

No context found.

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views After Redefinition. In SIGMOD, pages 211--222, 1995.

No context found.

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views after Redefinition. In Proceedings of ACM SIGMOD International Conference on Management of Data, pages 211--222, 1995.

No context found.

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views After Redefinition. In SIGMOD, pages 211--222, 1995.

No context found.

A. Gupta, I. Singh Mumick, and K. A. Ross. Adapting Materialized Views after Redefinitions. In SIGMOD, pages 211--222, 1995.

No context found.

A. Gupta, I. Mumick, and K. Ross. Adapting Materialized Views After Redefinition. In SIGMOD, pages 211--222, 1995.

No context found.

A. Gupta, I. S. Mumick, and K. A. Ross. Adapting Materialized Views After Redefinitions. In the ACM SIGMOD Conf., May 1995.

No context found.

A. Gupta, I. S. Mumick, and K. A. Ross. Adapting Materialized Views After Redefinitions. In the ACM SIGMOD Conf., May 1995.

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