A Model for Temporal Knowledge Visualization

以視覺化技術爲基礎之時間知識表達模式

侯建良(Jiang-Liang Hou) ； 陳育良(Yuh-Liang Chen)

《工業工程學刊》 24卷6期 (2007/11) Pp. 446-457

https://doi.org/10.29977/JCIIE.200711.0002

摘要

由於數位化資訊量膨脹迅速，為協助知識需求者快速且正確地吸收其所需知識與資訊，有效之知識表示方式已成為資訊提供者必須著眼探討之主題。而傳統知識分享環境中之文字型表達方式對於抽象化時間知識而言，難以讓讀者於短時間內精確掌握重要時間資訊；因此，若能藉由視覺化呈現方式表達對應之文字型時間知識，將可提升知識需求者對時間知識之理解與吸收速度。故本研究即針對時間相關知識提出一套「時間知識視覺化」方法論，以使電腦系統可將文字型時間知識內容自動轉為對應之視覺化圖形。本方法論之詳細作法可分為「全文標記」、「時序判斷」與「時間知識視覺化」等三大主軸，首先，藉由詞彙截斷與標記之方式擷取知識文件內容中所含時間相關資訊，並以之為基礎建立目標文句集合及產生對應事件內容，再透過時間相關資訊對各事件進行時序判斷，最後，將時間知識內容與圖形資料庫進行視覺化資料比對，並以圖形之表達方式將時間知識予以視覺化呈現。整體而言，本研究提出之時間知識視覺化方法論將可有效改善知識需求者理解與吸收時間知識之速率，並可延伸應用於業界以進行教育訓練與知識管理等活動。

關鍵字

知識表達 ； 時間知識 ； 視覺化 ； 知識管理

並列摘要

In order to assist knowledge receivers to efficiently and accurately understand the useful knowledge and information from a large number of digital documents, knowledge representation mechanism has become an important issue that the content providers should concern. In the traditional knowledge sharing environment, the abstract temporal knowledge such as temporal information or temporal relationship of events is hard to be fully recognized by knowledge receivers in a short time based on the text-oriented knowledge representation schemes. Therefore, a visual representation scheme for the text-oriented temporal knowledge will support knowledge receivers to efficiently recognize this type of knowledge. The core idea of this research is to extract and tag the temporal terms from text-oriented documents via temporal knowledge analysis and tagging procedure. As a result, the temporal relationships of events can be derived from the temporal information and the corresponding visualized display of temporal knowledge can also be established according to the extracted temporal information and events. This research develops a methodology for temporal knowledge visualization so that computer systems can automatically convert the text-oriented temporal knowledge into visualized display. A methodology for temporal information extraction and temporal relationship analysis is developed. Furthermore, a Web-based temporal knowledge extraction and representation system (TKERS) is also developed to demonstrate the feasibility of the proposed methodology. As a whole, this research provides an effective visualization methodology for the knowledge providers and receivers to improve efficiency for temporal knowledge recognition. The methodology can be further applied in many enterprise activities such as e-training or knowledge management to enhance reuse of temporal knowledge.

並列關鍵字

knowledge representation ； temporal knowledge ； knowledge visualization ； knowledge management

參考文獻

Aigner, W.,S. Miksch(2006).CareVis: Integrated visualization of computerized protocols and temporal patient data.Artificial Intelligence in Medicine.37,203-218.

Aigner, W.,S. Miksch,B. Thurnher,S. Biffl(2005).PlanningLines: Novel glyphs for representing temporal uncertainties and their evaluation.The 9th International Conference on Information Visualization.(The 9th International Conference on Information Visualization).:

Amghar, T.,D. Battistelli,T. Charnois(2002).Reasoning on aspectual-temporal information in French within conceptual graphs.Proceedings of the 14th IEEE International Conference on Tools with Artificial Intelligence.(Proceedings of the 14th IEEE International Conference on Tools with Artificial Intelligence).:

Berglund, A.,R. Johansson,P. Nugues(2006).A machine learning approach to extract temporal information from texts in Swedish and generate animated 3D scenes.The 11th Conference of the European Chapter of the Association for Computational Linguistics.(The 11th Conference of the European Chapter of the Association for Computational Linguistics).:

Cao, G.,W. Li,K. F. Wong,C. Yuan(2004).Combining linguistic features with weighted Bayesian classifier for temporal reference processing.Proceedings of the 20th International Conference on Computational Linguistics.(Proceedings of the 20th International Conference on Computational Linguistics).: