提升知識再利用效能之動態視覺知識呈現技術

Knowledge Reuse Enhancement with Motional Visual Representation

蔡宛蓉(Alice W.J. Tsai)

指導教授 ： 侯建良

國立清華大學/工學院/工業工程與工程管理學系/碩士(2007年)

https://doi.org/10.6843/NTHU.2007.00201

摘要

隨著科技技術快速發展、數位資訊大幅膨脹，資訊文件之類型亦日漸繁雜且多樣化，以致讀者不易於短時間內大量吸收。因此，為使讀者快速且準確地吸收其所需資訊與知識，良好的知識呈現方式乃成為現今資訊服務與知識管理之重要課題。在一般知識分享環境中，傳統以文字形式為主之知識表達形式易造成新進人員難以迅速吸收大量專業知識；而關聯性動作之相關知識因具有更多較為重要且操作較易混淆之句子，造成知識需求者往往需花費大量時間加以閱讀、理解。而因圖形化之知識表達方式可提升知識需求者學習速率與維持記憶之長久性等效果，本研究將針對關聯性動作相關知識，提出一套將關聯性動作相關知識制式化與視覺化之方法論，並根據該方法論建構一套可自動將文字資訊以視覺化方式動態呈現之系統。
本方法論之精神乃由自由形式知識文件中取得關聯性動作相關之文句，並將該等文句以標準化之制式形式表示，最後再將各制式形式文句以視覺化方式動態呈現。本方法論之詳細作法首先乃自動擷取自由形式文件中之新詞彙，並將其新增至詞庫中，以確保詞庫完整度。確保詞庫之完整性後，則將自由形式知識文件中各文句內容與關聯性動作詞庫進行比對，以取得隱含關聯性動作詞彙之目標文句，同時將該文句之關鍵詞彙予以解析並存入制式矩陣，以標準文句形式表示之。之後，將各制式矩陣中之關鍵詞彙與圖形庫進行比對，以取得對應之圖形與資訊，再以虛擬實境模組語言（VRML）呈現文句內容所對應之動態視覺效果。
最後，本研究更依據所提之方法論建構一套動態知識視覺呈現系統並以「電腦組裝」領域之文件資料進行案例驗證，以確認方法論與技術之可行性；而由驗證結果可知，本系統僅需一定數量之訓練資料即可達良好之績效。整體而言，本研究所提出之知識制式化與視覺化之方法將有助於知識需求者將大量資訊與知識快速且準確吸收，可大幅提升企業進行教育訓練與知識再運用之效率與效能。

關鍵字

知識表達 ； 動態知識 ； 視覺化 ； 知識管理 ； VRML

並列摘要

Due to the rapid growth of information technology, digital information has significantly increased over the Internet. The growing complexity of information and documents has made it hard for knowledge receivers to efficiently and accurately to recognize the digital contents. Therefore, an appropriate knowledge representation scheme is required for enterprise knowledge management and services. Traditional schemes for explicit knowledge representation within the enterprise and academic circles are mostly text-oriented and as a result, much time and efforts are required for knowledge receivers to recognize the knowledge contents, especially for the motion knowledge.
In this research, a three-phase methodology (including automatic thesaurus definition (ATD), target sentence extraction and formatting (TSEF) and motion knowledge visualization (MKV)) for motion knowledge extraction, representation and visualization is developed. Moreover, based on the proposed methodology, a Motion Knowledge Representation and Management System (MKRMS) is established, and a “Computer Assembly” case is applied in order to verify the feasibility of the proposed model. The verification results show that the system could achieve a well performance with simply a small amount of training data.
As a whole, this research provides a knowledge representation and visualization approach to facilitate knowledge receivers to efficiently and accurately acquire the knowledge contents. The proposed methodology can be applied in enterprise e-training and knowledge management systems to enhance reuse of domain knowledge.

並列關鍵字

Knowledge Representation ； Motion Knowledge ； Visualization ； Knowledge Management ； VRML

參考文獻

7. Baldwin, T. and Bond, F., 2003, “Learning the countability of English nouns from corpus data,” Proceedings of the 41st Annual Meeting of the Association of Computational Linguistics, pp. 463-470.

8. Boughanem, M., Layaida, R. and Caron, A., 1993, “A neural network model for documentary base self-organising and querying,” Proceedings of the 5th International Conference on Computing and Information, pp. 512-518.

9. Brown, M. E. and Gallimore, J. J., 1995, “Visualization of three-dimensional structure during computer-aided design,” International Journal of Human: Computer Interaction, Vol. 7, pp. 37-56.

10. Cammelli, A. and Socci, F., 1998, “A thesaurus for improving information retrieval in an integrated legal expert system,” Proceedings of the 9th International Workshop on Database and Expert Systems Applications, pp. 619-624.

11. Card, S. and Mackinlay, J., 1997, “The structure of the information visualization design space,” IEEE Computer Society Press, pp.92-101.