問答知識庫(Q&A Knowledge Base)在知識管理系統中扮演著儲存及提供重要知識的角色。然而目前的問答知識庫大多為領域知識專家所建構，因此常須耗費大量的建構成本。本論文旨在發展一個基於Ontology架構之Q&A系統，此系統可以自動的擷取及建構問答知識庫。在本論文中，我們提出一個包含有模糊推論與類神經網路的架構，並以此對文件進行問題與答案的萃取，我們使用平行模糊推論引擎來計算文件中各個句子的重要程度，再利用模糊推論推導各個句子關於Question的各種強度。而在類神經網路的部份，首先以倒傳遞類神經網路來判斷某一句子是否為一Question Sentence，再利用此類網路來判斷此一句子屬於何種問題類型。為了有效地擷取文件中的問題，我們使用自然語言處理技術來分析問題，並利用建構好的Ontology與知識庫來選取問題。最後再運用Ontology將擷取出的問題轉化成觀念化的特徵以利將來問答系統使用。由最後的實驗證實，此Q&A系統能夠有效地對中文文件進行問題擷取，來建構問答知識庫。

The Q&A knowledge base of an organization is important for knowledge management. It can provide experience and knowledge to a questioner. However, current Q&A knowledge bases are mostly constructed by domain knowledge experts, and the construction cost is expensive. This thesis proposes an ontology-based Q&A system that can construct Q&A knowledge base automatically. The ontology-based Q&A system extracts questions by using fuzzy inference engine and neural network. In the proposed architecture, the parallel fuzzy inference engine computes the important degree with sentences and infers the strength of each sentence for question property. In neural network model, the back-propagation learning algorithm is adopted to train the question extractor. In order to extract key questions in documents effectively, we apply natural language processing technology to analyze sentences of documents, and make use of many knowledge bases to aid the selection of questions. Finally, the extracted questions are conceptualized into the concepts of ontology, and could be search by a question answering subsystem. The experimental results exhibit that the proposed approach can extract questions for Chinese documents effectively.

[1] H. H. Chang, “Event Detection Driven Approach for Extracting Information from Internet Documents,” Master, Thesis, Department of Computer Science & Information Engineering, National Cheng Kung University, Taiwan, June, 2000.
[2] A. Preece, A. Flett, D. Sleeman, D. Curry, N. Meany, and P. Perry. “ Better knowledge management through knowledge engineering,” IEEE Intelligent System, vol.16, no. 1, Jan/Feb. 2001, pp. 36-43.
[3] R. Burke, K. Hammond, V. Kulyukin, S. Lytinen, N. Tomuro, and S. Schoenberg, “Question Answering from Frequently-asked Question Files:Experiences with the FAQ Finder System,” Technical Report TR-97-05, University of Chicage, Department of Computer Science, 1997.
[4] S. Kim, D. Baek, S. Kim, and H. Rim “Question Answering Considering Semantic Categories and Co-occurrence Density”, Proc. 9th Text Retrieval Conf.Maryland, Nov. 2000, pp.317-325.
[5] J. Prager, J. Chu-Carrol, and K. Czuba, “Use of WordNet Hypernyms for Answering What-Is Questions” , Proc. 10th Text Retrieval Conf.Maryland, Nov. 2001, pp.250-257.
[6] M. A. Pasca and S. M. Harabagiu, “High Performance Question/Answering”, Proc. 24th annual international ACM SIGIR conf. information retrieval, Sep. 2001, pp.336-374.
[7] K. C. Litkowski, “CL research experiments in TREC-10 question answering,” Proc. 10th Text Retrieval Conf.Maryland, Nov. 2001, pp.122-131.
[8] S. Vassiliadis, G. Triantafyllos, and W. Kobrosly, “A fuzzy reasoning database question answering system,” IEEE Transactions on knowledge and data engineering, vol. 6, no. 6 Dec. 1994, pp.868-882.
[9] H. Alani, S. Kim. D. Millard, M. Weal, W. Hall, P. Lewis, and N. Shadbot. “ Automatic ontology-based knowledge extraction from web documents,” IEEE Intelligent System, vol.18, no. 1, Jan/Feb. 2003, pp. 14-21.
[10] R. Navigli, P. Velardi, and A. Gangemi. “ Ontology learning and its application to automated terminology translation,” IEEE Intelligent System, vol.18, no. 1, Jan/Feb. 2003, pp. 22-31.
[11] R. N. Berthier, A. H. F. Laender, A. S. da Silva, “Ontology-based extraction and structuring of information from data-rich unstructured documents”, Proc. 8th CIKM Conf. Bethesda, MD, USA, Nov.1999, pp.52-59.
[12] S. G. Soderland, “CRYSTAL: Learning Domain-specific Text Analysis Rules,” CIIR Technical Report, 1996.
[13] C. Cardie, “Domain-specific Knowledge Acquisition for Conceptual Sentence Analysis,” PhD Thesis, CIIR Technical Report, 1994.
[14] E. M. Riloff, “Information Extraction as a Basis for Portable Text Classification Systems,” PhD Thesis, CIIR Technical Report, 1995.
[15] J. Kolodner, “Case-Based Reasoning,” Morgan Kaufmann, 1993.
[16] E. Brill, “ A Corpus-Based Approach to Language Learning,” PhD Dissertation, University of Pennsylvania, 1993.
[17] E. Brill, “Transformation-Based Error-Driven Learning and Natural Language Processing: A Case Study in Part of Speech Tagging Computational Linguistics,” 1995.
[18] E. Brill, “Unsupervised Learning of Disambiguation Rules for Part of Speech Tagging To appear in Natural Language Processing Using Very Large Corpora,” Kluwer Academic Press, 1997.
[19] J. F. Sowa, Knowledge Representation: Logical, Philosophical, and Computational Foundations, Pacific Grove, CA USA, 2000.
[20] N. Guarino, “Formal Ontology and Information System,” Proc. of the First International Conference (FOIS'98), Trento, Italy, June, 1998.
[21] N. Guarino. “The Role of Identity Conditions in Ontology Design,” Lecture Notes in Computer Science, Vol. 1661, pp. 221-234, 1999.
[22] “Academia Sinica Balanced Corpus,” Technical Report, No. 95-02/98-04, Academia Sinica, Taiwan, 1998.
[23] Y. H. Kuo, J. P. Hsu and C. W. Wang, “A Parallel Fuzzy Inference Model with Distributed Prediction Scheme for Reinforcement Learning,” IEEE Trans. on Systems, Man, and Cybernetics, Vol. 28, No. 2, pp.160-172, April, 1998.
[24] C. T. Lin, C. S. Gerorge Lee, “Neural-network-based fuzzy logic control and decision system,” IEEE Transactions on Computers, Vol. 40, No. 12, Dec. 1991.
[25] S. Haykin, “Neural Network,” Prentice-Hall, London, 1999.
[26] Head-Driven Principle, http://godel.iis.sinica.edu.tw/CKIP/treebank/newpage6.htm.