分散式物件整合資訊收集之研究與應用

Abstract

由於Web及Internet快速的擴增，有很多各種不同的資訊源(Information source) 被應用在上面，要在這種環境上找尋資訊變成一種困難的工作。一個整合式存取多個、分散式的、異質的、且獨立的資料庫或資訊源也因此是一個大型軟體發展重要的課題。目前電腦的研究及發展者面對著重要的挑戰是如何去發展一些容易整合及相互運作現有的資訊源的軟體。 在這篇論文中，我們要提出一個齊一的介面，這是一個在被認定為標準的分散式物件環境OMG的CORBA上提供一個資訊擷取及收集的介面，這齊一的介面將提供程式設計師一個擷取資訊的程式介面(API)，它可以讓應用程式擷取他所需要的資訊，我們也應用代理人技術來實作這個系統。另外，因為每一種型態的資訊來源都有它自己的查詢語言、綱目及屬性，在這個系統中是需要提供一個可擴充的環境來整合未來不同的資訊源。因此我們提出了一個可擴展的環境，它可以讓服務提供者在一個眾所週知的物件模式及語言下定義他們自已的資訊源的查詢語言及綱目。我們應用XML的DTD來定義資訊源的綱目，並且提供介面來管理它的中間資料(metadata)。委託者程式可以容易的啟動中間資料查詢的動作來取得資訊源的綱目。在論文中我們說明了設計及實作這個介面，並測量這個系統的效能，最後我們也應用這個系統實作了二個應用程式，分別是多搜尋引擎代理人及z39.50的包裝器。

With the advances of the Internet and the World Wide Web (for short WWW), there are more and more information published on it. Finding information on the Web and the Internet becomes a difficult task because of the Web’s tremendous size and a variety of kinds of information sources. An integrated access to multiple, distributed, heterogeneous, autonomous databases or information sources is therefore a significant topic in the development of large-scale software. In this dissertation, we are proposing a uniform interface for Information Retrieval and Gathering on the approved standard of distributed object-oriented environment, OMG’s CORBA. The unified interface will offer a programming interface for retrieving of what applications want, and applying agent technology to implement the infrastructure of IIG (Integrated Information Gathering). In addition, each type of information source has its own query language, schema, and attribute. It is necessary to support an extensible environment in this approach while integrating various information sources in the future. We want to propose an extensible environment, which can permit the source providers to define their own query interface and schema in a well-known object model and language. We apply the XML’s (eXtensible Markup Language) DTD to define the schema of information sources, and provide the interface in the IIG for managing metadata. The client is easily initiating the query operation of metadata to get the schema of information sources. Besides, we apply this technology to the applications of meta-search engine, Standardized Information Retrieval technology-Z39.50, etc.