鍾震
ZhenZhong
碩士論文(2012)
以需求導向知識獲取方法強化領域知識本體之研究
A Demand-Driven Knowledge Acquisition Method for Enhancing Domain Ontology Integrity
關鍵字 Keywords
需求、知識管理、知識本體、知識本體獲取、知識擷取
demand、knowledge management、ontology、ontology acquisition、knowledge retrieval
摘要
知識為現今經濟體系中最重要的資源,因此企業須有效地執行知識管理相關策略,在適當的時機提供正確的知識給對的使用者,以產生最高效益。而正確地表達知識為知識管理之基礎與成敗關鍵。知識本體結構化的知識表達模式有助於不同概念或語意間的轉換、交換與再利用,進而協助使用者更加流暢地運用知識,是目前最廣泛被接受的知識表達工具,然而快速成長的知識將導致領域知識本體完整性不足,並降低其使用價值。
本研究之目的為發展一強化領域知識本體之需求導向知識獲取方法,利用使用者之知識需求獲取領域知識本體所缺乏之知識概念,並與領域知識本體整合,以加強領域知識本體之完整性,進而提升其使用價值。為達上述研究目的,本研究主要研究項目包括(1) 強化領域知識本體之需求導向知識獲取流程設計,(2)需求前處理方法發展,(3)知識擷取與搜尋方法發展,(4)知識本體建構方法發展,(5)知識本體整合方法發展(6)強化領域知識本體之需求導向知識獲取機制實作。
Abstract
Knowledge has been the most important resource in the contemporary economic system. Enterprises need to take effective knowledge- management strategies to provide right knowledge to appropriate knowledge workers at a suitable time in order gain highest benefit. However, accurate knowledge representation is a fundamental and critical point for knowledge management among enterprises. Ontologies are the most popular and acceptable technology to represent domain knowledge due to its structurized representing fashion which performs well in semantic transition, transaction and reuse for knowledge concepts to the end of applying knowledge more smoothly by knowledge user. But the rapidly growth of knowledge with more and more interdisciplinary knowledge workers may relatively decrease the integrity of domain ontologies which reduces its value somehow.
This study proposed a Demand-Driven Knowledge Acquisition Method for enhancing the integrity of domain ontologies. This method acquires and integrates knowledge concepts which the original domain ontology lacked according to users’ knowledge demand in order to increases the value of domain ontologies. According to above mentioned purpose, the study first design a process model of “Demand-Driven Knowledge Acquisition for Enhancing Domain Ontology” and then develops following methods according to such model: (1) Demand Preprocessing, (2) Knowledge Retrieval and Searching, (3) Ontology Construction, (4) Ontology Integration. Finally, implement such model as a mechanism.