The distributed information search component (Disco) and the World Wide Web

Abstract

The Distributed Information Search COmponent (DISCO) is a prototype heterogeneous distributed database that accesses underlying data sources. The DISCO prototype currently focuses on three central research problems in the context of these systems. First, since the capabilities of each data source is different, transforming queries into subqueries on data source is difficult. We call this problem the weak data source problem. Second, since each data source performs operations in a generally unique way, the cost for performing an operation may vary radically from one wrapper to another. We call this problem the radical cost problem. Finally, existing systems behave rudely when attempting to access an unavailable data source. We call this problem the urzgrace~rd jadur-e problem.DISCO copes with these problems. For the weak data source problem, the database implementor defines precisely the capabilities of each data source. For the radical cost problem, the database implementor (optionally) defines cost information for some of the operations of a data source. The mediator uses this cost information to improve its cost model. To deal with ungraceful failures, queries return partial answers. A partial answer contains the part of the final answer to the query that was produced by the available data sources. The current working prototype of D tsco cent ains implementations of these solutions and operations over a collection of wrappers that access information both in files and on the World Wide Web.

Data integration using similarity joins and a word-based information representation language

“…In its basic motivation, our work is inspired by previous work in the integration of heterogeneous data sources, such as data sources on the Web [Levy et al 1996b;Arens et al 1996;Garcia-Molina et al 1995;Atzeni et al 1997;Tomasic et al 1997;Bayardo et al 1997]. None of these previous systems, however, include a "fuzzy" matching procedure for names; instead they construct global domains using hand-crafted domain-specific normalization schemes, or domain-specific matching algorithms [Fang et al 1994].…”

“…Integration of distributed, heterogeneous databases, sometimes known as data integration, is an active area of research in the database community [Duschka and Genesereth 1997b;Levy et al 1996b;Arens et al 1996;Garcia-Molina et al 1995;Tomasic et al 1997;Bayardo et al 1997]. Largely inspired by the proliferation of database-like sources on the World Wide Web, previous researchers have addressed a diverse set of problems, ranging from access to "semi-structured" information sources [Suciu 1996;Abiteboul and Vianu 1997;Suciu 1997] to combining databases with differing schemata [Levy et al 1996a;Duschka and Genesereth 1997a].…”

Working together in Harmony-an implementation of the CORBA object query service and its evaluation

“…The TSIMMIS project at Stanford addresses the problem of accessing non-standard data, notably semi-structured data, and proposes a flexible mediatorbased approach [3,8]. At INRIA, the Distributed Information Search Component (DISCO) has been developed [21,22]. However, all of these prototypes focus on heterogeneous query optimization and flexible data source integration using their proprietary middleware system.…”

Decoupled query optimization for federated database systems

“…Regardless of the number of cost dimensions, a centralized optimizer cannot accurately estimate the costs of operations at many autonomous sites. Garlic [23,40] and other middleware systems [24,46] address this problem by involving site-specific wrappers in the optimization process, but they do not consider the cost of communicating with these wrappers. This cost is not significant in these systems because the wrappers typically reside in the same address space as the optimizer.…”

“…The query optimization work goes back as far as the early distributed database systems (R*, SDD-1, Distributed Ingres [22,14,7]), and most recently has been focused on linking data sources of various capabilities and cost models [23,30,46]. However, query optimization in the broad federated environment presents peculiarities that change the trade-offs in the optimization process quite significantly.…”