J.A. Blakeley. Data Access for the Masses through OLE DB. SIGMOD Conference 1996.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... Optimizers The expansion of the Web has led to mediator prototypes that combine information from multiple heterogeneous sources [4, 9, 13, 18, 28, 37, 68, 71] Similarly, prototypes for integrating databases have been developed, and recently integration products are being released or announced [2, 5, 7, 26, 45]. We note that there is a close connection between mediator based systems [79] and distributed database systems [53] Many mediator systems have borrowed techniques for query optimization from the world of distributed databases. There has been a great deal of published work on these techniques ....

J. Blakeley. Data Access for the Masses through OLE DB. In Proc. ACM SIGMOD Conference, 1996.

....any of the movie review sites. Several systems have been built with the goal of answering queries using a multitude of Web sources [4,7,14,25,38,41,50,51,81,121] Many of the problems encountered in building these systems are similar to those addressed in building heterogeneous database systems [3,17,48,56,60,114,120]. Web data integration systems have, in addition, to deal with (1) large and evolving number of Web sources, 2) little meta data about the characteristics of the source, and (3) larger degree of source autonomy. There are two main differences between data integration systems and traditional ....

J.A. Blakeley, Data access for the masses through OLE DB, in: Proc. ACM SIGMOD Conference on Management of Data, Montreal, Quebec, 1996, pp. 161--172.

....and focus on relational projection, selection, and join. Our approach assumes that the underlying database is capable of most, or all, of the processing (though it may not be appropriate for the data source to actually evaluate the entire temporal query) Other mediator systems, e.g. OLE DB [Bla96] and DISCO [Tom97] are similar in their emphases on data integration and access to weak data sources (those with less powerful query facilities) We view mediator approaches as complementary to the temporal middleware approach introduced here. For integrating across diverse data sources, the ....

J. Blakeley. Data Access for the Masses Through OLE DB. In Proceedings of ACM SIGMOD, Montreal, Canada, pp. 161--172 (1996).

....IIS 9872128, and 9874759. 1 relevant to their query, interact with each source in isolation, and manually combine the data from the different sources. The problem of data integration has received significant attention in the research community as evidenced by numerous research projects (e.g. [18, 14, 44, 31, 17, 12, 4, 9, 2, 5, 1, 43, 25, 38, 16, 6, 23]) the emergence of several commercial products (e.g. DataJoiner [40] and companies providing data integration services (e.g. Netbot and Junglee) There are three main challenges distinguishing the design of a data integration system from that of a traditional database system: query ....

Jose A. Blakeley. Data access for the masses through OLE DB. In Proc. of ACM SIGMOD Conf. on Management of Data, pages 161--172, Montreal, Canada, 1996.

....manager, allows a shift from query shipping to data shipping, enabling operators that constitute the query to be executed on the client. This feature is extremely useful for building query services outside the kernel of a DBMS. In addition, a new branch of component based database research [2, 7, 6] and new approaches to supporting rules outside the kernel of database systems [8] which suggest factoring out from the DBMS kernel services that can be provided as complementary components, avoiding bloating the code image of the DBMS system when the functionality is not always required. ....

J. Blakeley. Data access for the masses through OLE DB. In Proc. of the ACM SIGMOD Intl. Conference on Management of Data, pages 161-172, 1996.

....the movie review sites. Several systems have been built with the goal of answering queries using a multitude of web sources [52, 42, 122, 82, 51, 38, 7, 25, 5, 14] Many of the problems encountered in building these systems are similar to those addressed in building heterogeneous database systems [3, 121, 61, 115, 49, 17, 57]. Web data integration systems have, in addition, to deal with (1) large and evolving number of web sources, 2) little meta data about the characteristics of the source, and (3) larger degree of source autonomy. There are two main di erences between data integration systems and traditional ....

Jose A. Blakeley. Data access for the masses through OLE DB. In Proc. of ACM SIGMOD Conf. on Management of Data, pages 161-172, Montreal, Canada, 1996.

....with speci c mostly passive database management systems [FGD98] Principally, new ECASs are provided by bundling them together out of existing, prefabricated components that realize speci c active database tasks. Hence, the FRAMBOISE approach follows a general direction as it is proposed in [Vas94, Bla96, GD94] and particularly in [GD98] namely to provide individual database management services that can be used and combined in a variety of ways and in a variety of environments. Obviously, there must be adequate components in order to build ECA Systems by putting together preexisting, reusable and ....

J. A. Blakeley. Data Access for the Masses through OLE DB. In Proc. SIGMOD, pages 161-172, 1996.

....data source. Lines represent exchanges of queries and answers. Solid boxes mark the application mediator protocol and the solid circles mark the mediator wrapper protocol. instead of the term database. A. Architecture As shown in Figure 1, current distributed heterogeneous database systems [1] [2], 3] 4] 5] 6] 7] 8] 9] 10] 11] adopt a distributed architecture that consists of several specialized components. End users interact with applications (A) written by application programmers. Applications access underlying data sources via mediators (M) 11] Mediators export a ....

....must be accepted by the wrapper. Higher conformance levels add additional features. Thus, at a minimum, the wrapper and the data source must support a variation of SQL. In Disco, a wrapper can support as little functionality as a simple file scan. In the abstract data type approach, e.g. 1] [2], the functionality of underlying sources is encoded as a black box function. This approach also simplifies the construction of mediators because, in general, the operations that can be pushed to a data source are fixed by the abstract data type interface. In [40] 41] 45] 30] to solve the ....

Jose Blakeley, "Data access for the masses through OLE DB," in Proceedings of theACM SIGMOD International Conference on Management of Data, New York, 1996, vol. 25, 2 of ACM SIGMOD Record, pp. 161--172, ACM Press.

....and tags used that are not mentioned in the DTD) the information is self describing. For structures that are more complex than flat tuples (e.g. in object relational systems) XML offers a way to serialize objects verbosely. Where efficiency is needed, other structures (e.g. OLE DB rowsets [BLAK96]) may prove more attractive. 4.2 XML for Document and Other Semi structured Data Relational DBMSs hold a small fraction of the world s data, for several good reasons. They tend to require professional administration. They require data to be tabular (i.e. flat) and to conform to a prespecified ....

J.A. Blakeley. "Data Access for the Masses through OLE DB," ACM-SIGMOD International Conference on the Management of Data, 1996, http://www.microsoft.com/data/oledb.

....using content and quality metadata. Acknowledgements: We are very grateful to Laura Bright and Tao Zhan for their research and programming support. 1 Introduction The rapid growth of the Internet and Intranets, vendor support of database interoperability protocols such as JDBC [26] OLE DB [6, 14], etc. and the emergence of XML to facilitate the exchange of semi structured data via the WWW, has dramatically increased the number of Web accessible data sources, WebSources. Wrapper mediator architectures [39] that are able to handle query processing with heterogeneous sources have been ....

J. Blakeley. Data access for the masses through ole db. Proceedings of the ACM SIGMOD Conference on Management of Data, 1999.

....Grants IRI 9303461, IIS 9872128, and 9874759. interact with each source independently, and manually combine the data from the different sources. The problem of data integration has received significant attention in the research community as evidenced by numerous research projects (e.g. [10, 20, 25, 17, 9, 3, 6, 1, 25, 21, 4, 13]) and the emergence of several commercial products (e.g. DataJoiner [23] and jango.excite.com) Three main challenges distinguish the design of a data integration system from that of a traditional database system: query reformulation, the construction of wrapper programs, and the design of new ....

J. A. Blakeley. Data access for the masses through OLE DB. In Proc. of ACM SIGMOD Conf. on Management of Data, pages 161--172, Montreal, Canada, 1996.

....is to benefit from database technology, without transferring the data into a DBMS. The approach consists in defining an open, extensible collection of interfaces that encapsulate orthogonal portions of DBMS technology. This set of interfaces, collectively referred to as OLE DB, is described in [Bla96] One key idea of this approach is the concept of component DBMSs, which consist of several independent data providing units, communicating with the main query processor through a set of well defined interfaces. This facilitates data integration while imposing minimal requirements on the ....

J. Blakeley. Data access for the masses through ole db. Proceedings of the ACM International Conference on Management of Data (SIGMOD), 1996.

....Research Project Agency under grant ARPA ONR 92 J1929, the National Science Foundation under grant IRI9630102 and by CONICIT, Venezuela. mediator (composition) query. The rapid growth of the Internet and Intranets, vendor support of database interoperability protocols such as JDBC [41] OLE DB [8, 28], etc. and the emergence of representations, e.g. XML Data, to facilitate the exchange of structured data via the WWW[49] has dramatically increased the number of Web accessible sources, WebSources. We propose WebSrcMed, a mediator for WebSources. Scaling to answer queries with a large number ....

Blakeley, J., "Data Access for the Masses through OLE DB", Proceedings of the ACM Sigmod International Conference, 1996.

....In other words, we unbundle active functionality from the DBMS. Thus, we follow a general direction that database research is currently about to take, namely to provide individual database management services that can be used and combined in a variety of ways and in a variety of environments [1, 2, 4, 11, 24]. We first discuss the advantages of unbundling active functionality. We then show how the unbundling process may take place. We start with a domain analysis of active DBMS style active functionality. The main task is the identification of components and the cooperation between them. This leads to ....

....and the cooperation between the units is not hidden within the system but can clearly be identified from outside. Considering database management systems as a form of traditionally large and monolithic software system, it makes sense to build them out from a number of cooperating components [1, 2, 4, 11, 24]. The major motivation for unbundling DBMS comes from the ever increasing demands on the functionality of DBMS because of (new) application areas (like data warehousing) new architectural forms (like distribution and heterogeneity) and technical innovations (like middleware) If one augments the ....

J. Blakeley. Data Access for The Masses through OLE DB. Proc. of the Intl. Conf. of the ACM SIGMOD, 1996.

....It also handles query decomposition into mediator subqueries, one for each source, and query optimization for these subqueries and the mediator (composition) query. The rapid growth of the Internet and Intranets, vendor support of database interoperability protocols such as JDBC [21] OLE DB [5, 6], etc. and the emergence of representations, e.g. XML Data, to facilitate the exchange of structured data via the WWW[18] has dramatically increased the number of available sources. Scaling to answer queries with a large number of data sources will have a disproportionate impact on the mediator ....

J. Blakeley. Data access for the masses through ole db. Proceedings of the ACM Sigmod International Conference, 1996.

.... cannot be migrated to a database, users would benefit from database services, such as query facilities or transactions in their tools [36] Thus, in the not too distant future we will see database systems expand and provide their services, e.g. to desktop applications, in a componentized form [4, 37]. Data stores managed by these tools and other data repositories will have to be integrated into large scale information systems, and database services must be available for them. These information systems will be operated over the Internet or Intranets, and users will query them, without willing ....

....on component technology. Wrappers are one possibility [29] where a wrapper encapsulates a specific data source and defines a predefined behavior, namely to understand queries and translate them into the query language of its data source. Another possibility is OLE COM, which is used in OLE DB [4] to define and implement components. Typical classes of components in OLE DB are service providers and data providers . A data Wrapper 3 Wrapper 2 Wrapper 1 Universal Query Service Data Source 2 Data Source 3 Data Source 1 Fig. 3. Universal Query Service provider is a component that can ....

J.A. Blakeley. Data Access for the Masses through OLE DB. In Proc. ACMSIGMOD Intl. Conf. on Management of Data, Montreal, Canada, June 1996.

.... using a multitude of web sources [GMPQ 97, EW94, WBJ 95, LRO96, FW97, DG97b, AKS96, Coh98, AAB 98, BEM 98] Many of the problems encountered in building these systems are similar to those addressed in building heterogeneous database systems [ACPS96, WAC 93, HZ96, TRV98, FRV96, Bla96, HKWY97] Web data integration systems have, in addition, to deal with (1) large and evolving number of web sources, 2) little meta data about the characteristics of the source, and (3) larger degree of source autonomy. An important distinction in building data integration systems, and therefore ....

Jose A. Blakeley. Data access for the masses through OLE DB. In Proc. of ACM SIGMOD Conf. on Management of Data, pages 161--172, Montreal, Canada, 1996.

....junyang db.stanford.edu middleware. Thus this first generation of middleware is not extensible to the arbitrary systems which may exist in a given business. Several projects are addressing the problem of middleware for increasingly diverse systems [SAD 94, PGMW95, TRV96b, ACPS96, LRO96, Bla96] Many of the data sources these systems integrate have limited or specialized query processing capabilities. One key challenge for these systems is to develop a general purpose query optimizer which can be told about the capabilities of a new data source, and which can use this information to ....

J. Blakeley. Data access for the masses through OLE DB. In Proc. of the ACM SIGMOD Conf. on Management of Data, pages 161--172, Montreal, Canada, June 1996.

....for those who build and manage desktop data in the Wintel world; all major vendors are working to integrate OLE support into their database engines and tools, and it is clear that support for OLE COM ADTs will be important in the future. Also looming on the horizon is Microsoft s OLE DB work [Blak96], which offers an approach to separating query optimization from execution in a world where data lives elsewhere in addition to databases. We will comment further on OLE DB in Section 4.2.4. Obviously, no discussion of object trends would be complete without touching upon the Java furor that has ....

....for data sources that yield relevance ranked results; and so on. This approach to legacy data is being explored in the Garlic project at the IBM Almaden Research Center [Care95] while semi structured data is a focus of the TSIMMIS project at Stanford [Garc95] Meanwhile, the Microsoft OLE DB work [Blak96] is also addressing this problem space, e.g. by defining the protocol for a query processor to use in talking to non database data sources. 4.2.5 Standards In the area of standards, SQL3 is moving in the direction that we have outlined, and it is drawing on OQL for inspiration in some areas. ....

J. Blakeley, "Data Access for the Masses through OLE DB," Proc. 1996 ACM SIGMOD Conference, Montreal, Canada, June 1996.

....It also handles query decomposition into mediator subqueries, one for each source, and query optimization for these subqueries and the mediator (composition) query. The rapid growth of the Internet and Intranets, vendor support of database interoperability protocols such as JDBC [33] OLE DB [5, 21], etc. and the emergence of representations, e.g. XML Data, to facilitate the exchange of structured data via the WWW[38] has dramatically increased the number of available sources. Scaling to answer queries with a large number of data sources will have a disproportionate impact on the mediator ....

Blakeley, J., "Data Access for the Masses through OLE DB", Proceedings of the ACM Sigmod International Conference, 1996.

....focus is on sources quite different from spreadsheets. Again, their abstraction and modeling of text essentially captures the structure of the text determined from a grammar for the language, rather than the application relevant semantics of the text. 4) Finally OLE and OLE DB have been claimed [4] to provide a database like view of Microsoft tools and are thus expected to enable programming interoperating applications among such sources in SQL. However, these technologies only achieve a syntactic database like view of the sources. The semantics are essentially left to the application ....

Blakeley, J.A. Data access for the masses through ole db. In ACM SIGMOD, pp161--172, 1996.

.... Optimizers The expansion of the Internet has led to mediator prototypes that combine information from multiple heterogeneous sources ( 3, 6, 14, 20, 22, 34] Similarly, prototypes for integrating databases have been developed, and recently integration products are being released or announced ([2, 4, 5, 18, 19, 24, 36]) We note that there is a close connection between mediator based systems ( 39] and distributed database systems. Given this, many mediator systems have incorporated query processing and optimization techniques of distributed databases. There has been a great deal of published work 1) X 1 : ....

J. Blakeley. Data Access for the Masses through OLE DB. In Proc. ACM SIGMOD Conf., pages 161--172, 1996.

.... 12, 18, 22, 23, 24] More recent work, based on an architecture of mediators and wrappers [27] and a common object model, is reported in [1, 7, 10, 11, 12, 17, 19, 25] The rapid growth of the Internet and Intranets, and vendor support of database interoperability protocols such as ODBC, OLE DB [5], and ODMG ODL, 8] etc. has increased the number of data sources accessible over networks. Scaling to a large number of data sources has also introduced several new problems. First, the Internet and Intranets are dynamic environments. One cannot guarantee that sources are always available. The ....

Blakeley, J. "Data Access for the Masses through OLE DB", Proceedings of the ACM Sigmod International Conference, 1996.

....support the directly specified queries. All capability extensions happen at the mediator level. The importance of information integration research is highlighted by the arrival of first generation commercial products. These include IBM s DataJoiner (Gupta and Lin [1994] and Microsoft s OLECOM (Blakeley [1996]) 8. Summary We hope the reader has obtained a feel for the principal innovations in the Tsimmis system: 4 Do not confuse the title, which begins with Compilers, with a subject, which does not exist in the figure. 1. The OEM data model, an object oriented model that uses object labels to ....

J. Blakeley [1996]. "Data Access for the Masses through OLE DB," ACM SIGMOD International Conf. on Management of Data, pp. 161-172.

....at a data site to support queries; this does not require importing the data into a full fledged DBMS, and existing applications on the data can run unchanged. We have developed such query handling software for use at a data site, along with a set of well defined interfaces that extend OLEDB [1]. Of course, where remote data sources provide query processing capabilities, the DEVise optimizer seeks to exploit this, like any distributed query optimizer. ffl Collaborative Data Analysis: DEVise enables several users to share visual presentations of the data, and to dynamically explore ....

J. A. Blakeley. Data access for the masses through oledb. In Proceedings of ACM SIGMOD, Montreal, Canada, 1996.

.... Existing Optimizers The expansion of the Internet has led to mediator prototypes that combine information from multiple heterogeneous sources ( 1, 10, 17, 23] Similarly, prototypes for integrating databases have been developed, and recently integration products are being released or announced ([2, 9, 13]) We note that there is a close connection between mediator based systems and distributed database systems. Given this, many mediator systems have incorporated query processing and optimization techniques of distributed databases. There has been a great deal of published work on these techniques ....

J. Blakeley. Data Access for the Masses through OLE DB. In Proc. ACM SIGMOD Conf., 161--172, 1996.

No context found.

J.A. Blakeley. Data Access for the Masses through OLE DB. SIGMOD Conference 1996.

No context found.

J. Blakeley. Data Access for the Masses Through OLE DB. In Proceedings of ACM SIGMOD, Montreal, Canada, pp. 161--172 (1996).

No context found.

Jos e A. Blakeley. Data access for the masses through OLE DB. In Proceedings of the 1996 ACM SIGMOD International Conference on Management of Data, Montreal, Quebec, Canada, June 4-6, 1996, pages 161--172, 1996.

No context found.

J. A. Blakeley. Data access for the masses through OLE DB. pages 161--172, Montreal, Canada, 1996.

No context found.

J. Blakeley, "Data Access for the Masses through OLE DB," Proceedings of ACM-SIGMOD, 1996.

No context found.

Jose Blakeley. Data Access for the Masses through OLE-DB. In Proceedings of ACM SIGMOD '96 International Conference on Management of Data, Montreal, Canada, pages 161--172, 1996.

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