L. Liu and C. Pu, "The Distributed Interoperable Ob- ject Model and its Application to Large-Scale Interoperable Database Systems", Fourth International Conference on Information and Knowledge Management, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....sources. The emphasis in the TSIMMIS system is that of automatic generation of translators and mediators for accessing and combining information in heterogeneous data sources. The resulting information is expressed in an Object Exchange Model. The Distributed Interoperable Object Model (DIOM) [25] shares similar goals with the TSIMMIS project. Also, similar runetionality can be found in lite resource agents in the Info Sleuth architecture. Additionally, we are considering ap proaches for automatic generation of translators and mediators for resource agents. In InfoSleuth, these are ....

L. Liu and C. Pu, "The Distributed Interoperable Ob- ject Model and its Application to Large-Scale Interoperable Database Systems", Fourth International Conference on Information and Knowledge Management, 1995.

.... the joint project Virtuelle Wissensfabrik (The Virtual Knowledge Factory) and resulted in the paradigm of mediated information systems [39, 41] Some current projects on mediation are TSIMMIS [37] HERMES [35] Information Manifold [21] SIMS [2] AURORA [42] DISCO [36] Squirrel [17] DIOM [23], Garlic [8] OBSERVER [25] InfoSleuth [3] and MMM [4, 5] In mediated information systems a client, seeking for information, and various and autonomous sources, holding potentially useful data, are brought together by a third kind of independent components, called mediators. Mediation is ....

L. Liu and C. Pu. Distributed interoperable object model and its application to large-scale interoperable database systems. In Proceedings of ACM International Conference on Information and Knowledge Management (CIKM'95), 1995.

....discussion should not be seen as a critique of these techniques. Rather, these techniques have been proposed primarily for data centric environments, where data is well organized and controlled. When applied to an open information universe as the Internet, these assumptions no longer hold (see [21] for a summary of desired system properties in the Internet) and some of the techniques do not easily extend to scale up to the distributed interoperable environment. 7.2.1 Active Databases Most of active database systems [38] provide facilities [6, 30, 35] that allow users to specify, in the ....

L. Liu and C. Pu. The distributed interoperable object model and its application to largescale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....provide raw data, wrappers that provide interfaces to data sources, mediators that provide declarative query access to multiple wrappers, and clients that provide queries to mediators and accept answers returned from mediators. Several projects follow a similar architecture (e.g. Garlic [3] DIOM [4], HERMES [5] COIN [6] IRO DB [7] or related architectural frameworks. Declarative access in the form of queries on data sources gives a degree of freedom to the mediator to determine the best plan for the execution of the query. From a declarative query, the mediator can generate multiple access ....

Ling Liu and Calton Pu, "Distributed Interoperable Object Model and Its Application to Large-scale Interoperable Database Systems", in 4th Int'l Conf. on Information and Knowledge Management (CIKM), Baltimore, Maryland, 1995.

....these multiple heterogeneous sources. The Rainbow system consists of several components that gather information from unstructured data sources or from other information brokers mediators, and represent gathered information in terms of a consumer s interface description defined using DIOM IDL IQL [32]. The main component services that the RAINBOW prototype addresses include the producers information source registration, metadata library management, query routing, query decomposition, parallel query plan generation, subquery transformation and execution, and results assembly. This thesis will ....

....on a network of simpler mediators which interact with each other in order to accomplish tasks. For example, a Document Inquiry mediator (as shown in Figure 2. 2) can be built by interacting among several smaller mediators such as Book mediator, TechReport mediator, Medical ClaimFolder mediator [32], and so on. To build a network of specialized and interacting mediators, we need an architecture for a single mediator (perhaps called a meta mediator) that can be instantiated to provide multiple mediators. The ultimate goal of developing such a meta mediator architecture is to provide a uniform ....

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....that provide interfaces to data sources, mediators that provide declarative query access to multiple wrappers, and clients that provide queries to mediators and accept answers returned from mediators. Several projects follow a similar architecture (TSIMMIS [GMHI 94] Garlic [C 95] DIOM [LP95] Information Manifold [KLSS95] HERMES [SAB 97] COIN [GMS94] IRO DB [G 95] or related architectural frameworks (Infosleuth [R 96] Declarative access in the form of queries on data sources gives a degree of freedom to the mediator to determine the best plan for the execution of ....

Ling Liu and Calton Pu. Distributed Interoperable Object Model and Its Application to Large-scale Interoperable Database Systems. In 4th Int. Conf. on Information and Knowledge Management (CIKM), Baltimore, Maryland, 1995.

....networks. The canonical model is often an object oriented one [14] It has been recognised that such an environment requires flexible and scalable architectures, where users of the component databases are provided with tools to establish importation of information from foreign data sources [9,11,23]. Tightly coupled approaches, where the schemata of all component databases are unified into a global schema by a central modelling authority possessing a helicopter view of all component databases, are generally agreed to be infeasible in such situations, if only because the component databases ....

L. Liu & C. Pu, "The Distributed Interoperable Object Model and its application to large-scale interoperable database systems," in Proceedings Fourth International Conference on Information and Knowledge Management (CIKM), Baltimore, Maryland, 1995 .

....that are crucial for the user s query. Wrappers can be realized using distributed object managers such as CORBA [25] or software agents [14, 1] Some current projects on mediation are TSIMMIS [35] HERMES [33] Information Manifold [20] SIMS [4] AURORA [40] DISCO [34] Squirrel [16] DIOM [22], Garlic [10] OBSERVER [24] InfoSleuth [5] and MMM [7, 8, 2] There also appeared some contributions to security in mediated information systems [9, 39, 18] The mediators may use dioeerent approaches to support the schema, e.g. materialized, virtual, and hybrid [8] The materialized approach ....

L. Liu and C. Pu. Distributed interoperable object model and its application to large-scale interoperable database systems. In Proceedings of ACM International Conference on Information and Knowledge Management (CIKM'95), 1995.

....de Donnes htrognes distribues, traitement de requtes, valuation partielle, donnes indisponibles 3 1 Introduction Many current application environments provide declarative access to a wide variety of heterogeneous data sources. Research into improving these systems has produced many new results [8, 7, 17, 1, 4, 12, 13, 3, 15, 11], which are being incorporated into prototypes and commercial products. However, to the best of our knowledge, these systems all fail ungraciously in the presence of unavailable data sources. They either assume that all data sources are available, report error conditions, or silently ignore ....

Ling Liu and Calton Pu. Distributed interoperable object model and its application to large-scale interoperable database systems. In Fourth International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, 1995.

....operation automatically begins a nested transaction so that the calling operation is insulated from network failures. Due to the space limitation, we omit the discussion on the DIOM built in type hierarchy in this paper. Readers who are interested in this part may refer to our technical report [16]. 10 3.3 Advanced Concepts for Interface Definitions 3.3.1 Base Interface and Compound Interfaces As mentioned earlier, a type of distributed interoperable objects is defined by specifying its interface in the DION Interface Definition Language (DION IDL) The DION interfaces are classified ....

....mechanism. Figure 5 shows how each of the base interfaces Claim, Image, and Document is specialized by adding the required relationships into its specialized interface definition in DIOM IDL. Due to the space limit, we omit the the BNF syntax of the DIOM IDL in this paper. Readers may refer to [16] for detail. 3.3.3 Interface Abstraction Mechanism: Generalization The generalization abstraction provides a convenient facility to merge several semantically similar and yet different interfaces into a more generalized interface. The main idea is based on generalization by abstracting the ....

[Article contains additional citation context not shown here]

L. Liu and C. Pu. The distributed interoperable object model and its application to large scale interoperable database systems. Technical report, University of Alberta, Feb. 1995.

....a nested transaction so that the calling operation is insulated from network failures. Due to the space limitation, we omit the discussion on the DIeM built in type hierarchy and the BNF syntax of the IDL in this paper. Readers who are interested in this part may refer to our technical 10 report [20]. 3.3 Advanced Concepts for Interface Definitions 3.3.1 Base Interface and Compound Interfaces As mentioned earlier, a type of distributed interoperable objects is defined by specifying its interface in the DION Interface Definition Language (DION IDL) The DION interfaces are classified into ....

....can automatically be propagated to the subtypes that are specialized versions of it. In the first phase of DIOM implementation, the specialization abstraction is only supported for construction of a new interface based on one or more base interfaces whose scope is the same data repository (see [20] for more detail) 3.3.5 The Import Mechanism The import mechanism is designed for importing selected portions of the data from a given export schema, instead of importing everything that is available. For a data repository that manages complex objects, the import mechanism performs the ....

[Article contains additional citation context not shown here]

L. Liu and C. Pu. The distributed interoperable object model and its application to large scale interoperable database systems. Technical report, University of Alberta, Feb. 1995.

....influenced by the work in active databases and view materialization. Both areas have targeted primarily data centric environments, where data is well structured, well organized and controlled. When applied to an open information universe as the Web, many of these assumptions no longer hold (see [15] for a summary of desired system properties in the Internet) Conquer bridges this gap. Active Databases: Event Condition Action (ECA) systems are rule based programs in which an event triggers the testing of a condition, which in turn (if true) triggers an action. Most of active database systems ....

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995. 23

....influenced by the work in active databases and view materialization. Both areas have primarily for data centric environments, where data is well structured, well organized and controlled. When applied to an open information 20 universe as the Web, many of the assumptions no longer hold (see [15] for a summary of desired system properties in the Internet) Conquer bridges this gap. Active Databases: Event Condition Action (ECA) systems are rule based programs in which an event triggers the testing of a condition, which in turn (if true) triggers an action. Most of active database systems ....

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....GSB 615, Edmonton, Alberta T6G 2H1 Canada email: fyoosh,lingliug cs.ualberta.ca Calton Pu y Dept. of Computer Science and Engineering Oregon Graduate Institute P.O. Box 91000 Portland, Oregon 97291 1000 USA email: calton cse.ogi.edu Abstract The Distributed Interoperable Object Model (DIOM) [19, 17] introduced the approach that explicitly defines the interfaces of an information consumer and an information producer, matching them dynamically to achieve interoperability in heterogeneous information systems with growing number of autonomous data sources as components. In this paper, we ....

....in building an experimental cooperative information system, called Diorama, capable of providing transparent and customizable information access across multiple heterogeneous and autonomous information sources. The theoretical model and architecture of Diorama are based on previous work on DIOM [17, 19, 21], which features the support of the USECA properties throughout the development of distributed query services. 1 Diorama demonstrates the systematic design and the flexible cooperation of the DIOM distributed query mediation services, and the practical applicability of the DIOM approach [19, 21] ....

[Article contains additional citation context not shown here]

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....redundant or nearly redundant rules. As pointed out in [6, 10] large amount of rules could potentially be pruned if there were adequate ways to remove redundancy. We are also interested in applying data mining techniques to resource discovery in global and interoperable information systems [14, 13]. ....

Ling Liu and Calton Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....a nested transaction so that the calling operation is insulated from network failures. Due to the space limitation, we omit the discussion on the DIOM built in type hierarchy and the BNF syntax of the IDL in this paper. Readers who are interested in this part may refer to our technical report [20]. 3.3 Advanced Concepts for Interface Definitions 3.3.1 Base Interface and Compound Interfaces As mentioned earlier, a type of distributed interoperable objects is defined by specifying its interface in the DIOM Interface Definition Language (DIOM IDL) The DIOM interfaces are classified into ....

....can automatically be propagated to the subtypes that are specialized versions of it. In the first phase of DIOM implementation, the specialization abstraction is only supported for construction of a new interface based on one or more base interfaces whose scope is the same data repository (see [20] for more detail) 3.3.5 The Import Mechanism The import mechanism is designed for importing selected portions of the data from a given export schema, instead of importing everything that is available. For a data repository that manages complex objects, the import mechanism performs the automatic ....

[Article contains additional citation context not shown here]

L. Liu and C. Pu. The distributed interoperable object model and its application to large scale interoperable database systems. Technical report, University of Alberta, Feb. 1995.

....of these database techniques with each other and with the relatively unorganized storage and management of data, such as WWW pages [4] is completely non trivial. To address the interoperability problem in Internet environments we have developed the Distributed Interoperable Object Model (DIOM) [10], which supports a methodical and explicit composition of query results from interoperable information sources. The key idea of this paper is to generate new query results by incrementally updating previous query results, rather than re evaluating the queries from scratch. This approach builds on ....

....of query results from interoperable information sources. The key idea of this paper is to generate new query results by incrementally updating previous query results, rather than re evaluating the queries from scratch. This approach builds on our previous results from the Diorama architecture [10, 11] and Epsilon Serializability [24, 17] to support flexible and efficient monitoring of information updates in the Internet environments. The first contribution of this paper is our use of DIOM and Epsilon Serializability to define a powerful mechanism for user specification of update monitoring. ....

[Article contains additional citation context not shown here]

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....become accessible, including multimedia data and web pages. Some of the fundamental assumptions in traditional databases, such as the existence of a global schema and data consistency maintained by a Data Base Administrator, are no longer true in many of the new data sources. We outline the DIOM [LP95b] object oriented approach to build interoperable heterogeneous information systems despite the absence of global schema and the presence of data inconsistency. We describe the metadata catalog management component of DIOM, a key service in the support for interoperation among heterogeneous ....

....they are defined as programs that take a database from a consistent state to another consistent state. In the current generation of new information systems (or so called non traditional database systems (NDBs) such as scientific data [BP93] multimedia databases, a variety of Internet data sources [LP95b], some of these important assumptions do not hold and classic database techniques may not apply directly. In this paper, we examine two decreasingly valid assumptions in nontraditional databases that strive to be interoperable: 1) data is consistent in the database, and (2) database schema is ....

[Article contains additional citation context not shown here]

Ling Liu and Calton Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....a nested transaction so that the calling operation is insulated from network failures. Due to the space limitation, we omit the discussion on the DIOM built in type hierarchy and the BNF syntax of the IDL in this paper. Readers who are interested in this part may refer to our technical report [20]. 3.3 Advanced Concepts for Interface Definitions 3.3.1 Base Interface and Compound Interfaces As mentioned earlier, a type of distributed interoperable objects is defined by specifying its interface in the DIOM Interface Definition Language (DIOM IDL) The DIOM interfaces are classified into two ....

....can automatically be propagated to the subtypes that are specialized versions of it. In the first phase of DIOM implementation, the specialization abstraction is only supported for construction of a new interface based on one or more base interfaces whose scope is the same data repository (see [20] for more detail) 3.3.5 Interface Composition Mechanism: Import The import mechanism is designed for importing selected portions of the data from a given export schema, instead of importing everything that is available. For a data repository that manages complex objects, the import mechanism ....

[Article contains additional citation context not shown here]

L. Liu and C. Pu. The distributed interoperable object model and its application to large scale interoperable database systems. Technical report, University of Alberta, June 1995.

....consumers query requests, and (2) scalability of distributed query services in the presence of a growing number of information sources and the evolving requirements of both information producers and information consumers. We have proposed the Distributed Interoperable Object Model (DIOM) [10] in the context of the Diorama project as a concrete mediator based approach to interoperability through the cooperation among a network of specialized mediators. Our early work mainly focuses on the design of the DIOM object model which captures the requirements of advanced distributed ....

....as a concrete mediator based approach to interoperability through the cooperation among a network of specialized mediators. Our early work mainly focuses on the design of the DIOM object model which captures the requirements of advanced distributed information systems as the USECA properties [10]: Uniform access to heterogeneous multimedia information sources, Scalability to the growing number of information sources, Evolution and Composability of software and information sources, and Autonomy of participants, both information consumers and information producers. In this paper we ....

[Article contains additional citation context not shown here]

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....an open environment. 3 Background and Terminology 3. 1 An Overview of the DIOM system The DIOM [26] system has two tier architecture and offers services at both the mediator tier and the wrapper tier (see Figure 1) Mediators are software modules that handle application specific service requests [25]. One of the main tasks of the mediator sub system is to utilize the metadata obtained from both information consumers (i.e. user profiles) and information producers (i.e. source capability descriptions) for efficient processing of distributed queries [24] Wrappers are software modules that ....

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

....these techniques have been proposed primarily for data centric environments, where data is well organized and controlled. When applied to an Information Superhighway environment (for brevity and concreteness we refer to this environment as the Internet) these assumptions no longer hold (see [13] for a summary of desired system properties in the Internet) Therefore, we discuss these assumptions before we launch into our definition of continual queries (Section 3) and a technique (Section 5) for efficient processing of continual queries. 2.1 Continuous Queries Terry et al. [27] proposed ....

....delta relations. We have proved that, for any SPJ expression, using differential re evaluation method is functionally equivalent to the complete re evaluation of the query. In this section we will discuss a number of issues related to the implementation of continual queries in the Diorama project [13]. 6.1 Improving Performance in Compilation and Execution Our differential algorithm for query refresh does not automatically provide the most efficient way of processing CQs. There are a number of possibilities to further improve the performance of differential processing of continual queries. ....

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. Technical Report OGI-CSE-95-003, Department of Computer Science and Engineering, Oregon Graduate Institute, February 1995.

....partial matching and correlations, and other deep information semantics. In this paper, we focus our attention on the first two issues. We make two contributions towards flexible and efficient monitoring of information updates, motivated by a distributed interoperable environment such as DIOM [12]. First, we describe continual queries with epsilon specification, namely epsilon continual queries (ECQ) to simplify continual query formulation on the user side. In addition to triggering query refresh periodically as proposed in previous work [27, 2] ECQs allow the users to specify and ....

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. Technical report, Portland, OR, March 1995.

.... and information producers, providing transparent and customizable information access across multiple heterogeneous and autonomous information sources (including databases, knowledge bases, flat files, or programs) We have captured the requirements of these systems as the USECA properties [14]: Uniform access to heterogeneous information sources, Scalability to the growing number of information sources, Evolution and Composability of software and information sources, and Autonomy of participants, both information consumers and information producers. We have proposed the Distributed ....

....information sources, Scalability to the growing number of information sources, Evolution and Composability of software and information sources, and Autonomy of participants, both information consumers and information producers. We have proposed the Distributed Interoperable Object Model (DIOM) [14] in the context of the Diorama project as a concrete information mediation architecture to support the cooperation in a network of specialized mediators. The distinctive features of the DIOM architecture include: the support for USECA properties in the development of distributed object query ....

[Article contains additional citation context not shown here]

L. Liu and C. Pu. The distributed interoperable object model and its application to large-scale interoperable database systems. In ACM International Conference on Information and Knowledge Management (CIKM'95), Baltimore, Maryland, USA, November 1995.

No context found.

Liu, L.; Pu, C. (1995). Distributed Interoperable Object Model and Its Application to Large-scale Interoperable Database Systems. In: Proceedings of ACM International Conference on Information and Knowledge Management (CIKM'95).

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