以XML原生資料庫實現FAQ知識查詢系統之研究
Abstract
隨著網路的快速發展，許多人遇到問題時，往往透過搜尋Google、維基百科等入口網站來尋求解答。有些網站，例如銀行、政府網站等，會利用常見問題(FAQ)來提供特定範圍的知識，讓使用者更快速的取得相關知識。然而這些FAQ資訊往往以靜態網頁的方式來呈現，只適合少量的問題，且無法做有效的管理，因此如何妥善管理及規劃特定領域的知識是本研究的重要探討議題。
本研究以IPv6相關知識為例，來探討如何建立多維度有結構性的FAQ知識查詢系統。由於傳統的FAQ通常只有問題與答案的文字內容，一個完整的知識內容必須具有豐富的內容，所以我們利用XML檔案具有半結構化的優點，將每一個FAQ存成XML檔案。為了保留XML文件本身結構的優點，我們使用XML原生資料庫(eXist)來儲存FAQ的檔案，透過XQuery查詢語言，可以對XML檔案的結構作結構性的查詢。而當知識量越來越多的時候，傳統的搜尋方法往往會讓使用者找到許多不相關的答案，使用者需要花費時間來篩選正確的答案。所以我們使用OWL/RDF將XML原生資料庫中的FAQ檔案做樹狀式的儲存，讓使用者能透過多維度的概念，可以便利地執行FAQ查詢。除此之外，為了能夠讓使用者更加了解所要的知識輪廓，我們應用知識概念圖的概念允許使用者利用知識概念圖來作FAQ的搜尋。本研究提出的方法有別於傳統的FAQ系統，其結合了語意網路相關技術，提供一套擁有管理、上傳及多維度查詢功能的FAQ知識查詢系統，讓管理者可以更有效率的管理FAQ，同時也讓使用者能夠更精確的找尋所需要的答案。
Nowadays, Internet has become the major information source for many people. When people need answer for certain problems, they often search Google, Wikipedia or other portals for the answers. Many websites, such as banks and government, use Frequently Asked Questions (FAQs) to provide a specific range of knowledge, which provides users fast access to relevant knowledge. However, traditional FAQ service is usually provided by static webpages which has many limitations. Therefore, how to properly manage and plan FAQs for some specific domain knowledge is an important issue. Our research takes IPv6 related knowledge as an example to explore how to establish a structured multi-dimensional FAQ Knowledge Query System. For flexibility, we use the native XML-based database(eXist) to store these XML files. On the other hand, we proposed a method using RDF/OWL documents to depict the classifications, and allow users to search from multiple dimensions. Furthermore, we provide concept map FAQ search to assist users in comprehending the searched knowledge. This research combines Semantic Web technologies to enhance the efficiency of the management of FAQs and allow users to obtain the information more accurately.
Nowadays, Internet has become the major information source for many people. When people need answer for certain problems, they often search Google, Wikipedia or other portals for the answers. Many websites, such as banks and government, use Frequently Asked Questions (FAQs) to provide a specific range of knowledge, which provides users fast access to relevant knowledge. However, traditional FAQ service is usually provided by static webpages which has many limitations. Therefore, how to properly manage and plan FAQs for some specific domain knowledge is an important issue. Our research takes IPv6 related knowledge as an example to explore how to establish a structured multi-dimensional FAQ Knowledge Query System. For flexibility, we use the native XML-based database(eXist) to store these XML files. On the other hand, we proposed a method using RDF/OWL documents to depict the classifications, and allow users to search from multiple dimensions. Furthermore, we provide concept map FAQ search to assist users in comprehending the searched knowledge. This research combines Semantic Web technologies to enhance the efficiency of the management of FAQs and allow users to obtain the information more accurately.
Description
Keywords
Protégé, eXist資料庫, 資訊網本體論語言, XQuery, Protégé, eXist Database, OWL, RDF, XQuery