在數位學習環境中，學習者比較容易缺乏同儕與教師間的互動，當遇到課程相關的問題時，往往得不到即時的解決，必須透過線上即時問答系統，或者透過留言版或電子郵件來提出問題；但這必須得仰賴老師或者教學助教即時的在線上提供回應，學習者才能得到解答。如何提供自動化的學習輔助及自我學習機制，則是目前數位學習環境的一個重要發展目標；因此如何設計一個能提供自動化即時回答與教材相關的答問機制，對於學習者的學習成就將會相當助益。
因此本論文提出一個支援語意認知方法的答問系統（Semantic QA），採用知識本體技術設計資料結構課程教材，以期能切要回答學習者以自然語言所提問之課程知識。
根據上述目標，本論文針對語意分析所採用相關研究方法為：（1）使用Link Grammar Parser來分析學習者所提之問句句法資訊建構句法樹，（2）分析問句類型取得標的片語並建構一個語意樹以表示問句句意，（3）藉由WordNet取得相似字清單以得到更相關的語意資訊，（4）最後將語意樹與課程知識本體比對取得學習者所提問題之答案。
而針對於知識本體內容完整性，本論文設計提供知識本體延伸模組分析網路上相關主題之教材，以豐富答問系統知識本體；此外並提供教師回饋機制，教師可藉由此模組觀察系統是否可正確回答學習者問題，並可補充答問系統相關之複合字、問句類型、相似字清單等問句分析的相關資訊。
本論文發展的問答系統具備以下特性：（1）瞭解發問者以自然語言形式所提出的問題，（2）增加回答問題的準確率，（3）可以延伸問句分析的相關資訊，(4) 建立自動化的問答系統以輔助學習。

Different from those in the traditional learning, learners in the e-learning environment usually have less opportunity to interact with their teachers or fellows. If learners have questions, they have to seek for answers through the on-line question answering system or post their questions on message board or use e-mail to ask questions. However, teachers or tutors can not solve each question immediately if they are not online simultaneously. Therefore, how to build an automatic assistance for learning and provide a self-paced learning mechanism are the objectives in today’s e-learning environment. For this reason, constructing an automatic question answering mechanism will be very helpful for learners to get solutions instantly.
According to the above motivation, this thesis proposes a Semantic-extended Question Answering system (Semantic QA) to analyze learners’ questions and find the relevant answer from the target course ontology which applies the Data Structure to construct teaching materials. 
To achieve this goal, the following shows the research steps. Firstly, Link Grammar Parser is applied to analyze the Syntactic Information of an input sentence and form a syntactical tree. Secondly, the question type is analyzed to find the target phrase and then form a Semantic Tree to represent the meaning of the question. Thirdly, WordNet is used at this stage to generate the Similar Word Lists to extend relevant meaning. Lastly, the Semantic Tree will be mapped to the Data Structure course ontology and find the relevant contents in order to answer them to learners.
The Semantic QA system has the following characteristics: (1) understand learners’  questions in the form of natural language, (2) enhance the accuracy of solutions, (3) make the extensibility of the related question analysis information and (4) establish the automatic learning mechanism.

Contents

1	Introduction	1
2	Theoretical Background	5
2.1	Link Grammar	5
2.2	WordNet	12
2.3   Ontology	13
3	System Architecture	18
3.1	Background Research	18
3.2	System Architecture Design	20
3.2.1	User Interface:	22
3.2.2	Syntactic Analysis:	23
3.2.3	Semantic Analysis:	24
3.2.4	Ontology Design and Extension:	24
4	System Design	25
4.1	Syntax Analysis	27
4.1.1	Replace Compound Word	29
4.1.2	Get Syntactic Information	31
4.1.3	Restore Compound Word	32
4.1.4	Construct Syntactical Tree	32
4.2	Semantic Analysis	34
4.2.1	Question Sentence Pattern Match	34
4.2.2	Semantic Tree Construction	37
4.2.3	Similar Word Sense Extension	40
4.2.4	Map the Data Structure Course Ontology	44
4.3	Ontology Design and Extension	48
4.3.1	Data Structure Course Ontology Design	48
4.3.2	Build Domain Terminology	51
4.3.3	Ontology Extension Module	52
4.4	Instructor Feedback for System	55
5	Evaluation	57
6	Conclusions and Future Works	61
7	Bibliography	64
8	Appendix A. Publication List	67
9	Appendix B.	71
10	Appendix C.	93



 
List of Figures
Figure 2 1 Words and connectors in a dictionary	6
Figure 2-2. All linking requirements are satisfied	6
Figure 2 3. A simplified form of Figure. 2	7
Figure 2 4. The link grammar output Linkage example 1	10
Figure 2 5. The link grammar output Linkage example 2	10
Figure 2 6. The Link Grammar output sample through Phrase Parser component -1	10
Figure 2 7. The Link Grammar output sample through Phrase Parser component -2	11
Figure 2 8. An example of Ontology using KAON	17
Figure 3 1. System Functional Blocks	19
Figure 3 2. System Architecture	21
Figure 3 3. User Interface Layer	22
Figure 4 1. System Process Flow	26
Figure 4 2. Syntactical Analysis Layer	29
Figure 4 3. Incomplete Linkage example with “fist in first out”	30
Figure 4 4. Complete Linkage with “fist-in-first-out”	31
Figure 4 5. The output syntactic information with “constituents =1”	31
Figure 4 6. The output syntactic information with “constituents =2”	31
Figure 4 7. The XML form of the syntactic information	32
Figure 4 8. The Syntactical Tree of the example question sentence	33
Figure 4 9. The sample of retrieved target phrase -1	35
Figure 4 10. The sample of retrieved target phrase -2	35
Figure 4 11. Semantic Linking Unit	37
Figure 4 12. The Node presentation	38
Figure 4 13. The nodes dependency representation	38
Figure 4 14. The Semantic Tree example	39
Figure 4 15. Add similar words to Semantic Tree	43
Figure 4 16. The Refined Semantic Tree	46
Figure 4 17. The Partially Mapped Answers	47
Figure 4 18. Concept Layer of the Data Structure Course Ontology	50
Figure 4 19. Instance layer example of the Data Structure Course Ontology	51
Figure 4 20. Ontology Extension Module	53
Figure 4 21. Instructor Feedback for System	55

 
List of Tables

Table  2 1. The words and linking requirements in a dictionary	7
Table  2 2. WordNet Synset sample of “Add”	12
Table  4 1. Compound Words List	30
Table  4 2. Question Sentence Pattern Matching Table	34
Table  4 3. Similar Word List	40
Table  5 1. Evaluation with other famous systems	60

[ADL 2007] Advanced Distributed Learning (ADL), “Sharable Content Object Reference Model (SCORM),” [Online]. Available: http://www.adlnet.gov/, 2007
[Allen 1995] Allen, J., “Natural Language Understanding,” the Benjamin/Cummings Publishing Company, 1995 
[Bozsak 2002] Bozsak E., Ehrig M., Handschuh S., Hotho A., Maedche A. and Motik B., et al., ”KAON - Towards a large scale Semantic Web,” Third International Conference of E-Commerce and Web Technologies (EC-Web 2002), Aix-en-Provence, France, Vol. 2455 of LNCS, Springer, 2002, pp.304-313 
[DAML-S 2003] DAML-S SERVICES COALITION, “DAML-S: Semantic Markup for Web Service,” W3C Recommendation, [Online]. Available: http://www.daml.org/services/daml-s/0.9/daml-s.pdf, 2003
[Denny 2004] Denny, M., “Ontology Tools Survey, Revisited,” [Online]. Available:  http://www.xml.com/pub/a/2004/07/14/onto.html, 2004
[Didion 2003] Didion, J., “Java WordNet Library (JWNL 1.3)”, [Online]. Available: http://jwordnet.sourceforge.net/, 2003
[IAC 2007] IAC Search & Media., “ASK.com,” Oakland, California, [Online]. Available: http://ask.com/, 2007
[Jordan 2007] Jordan C., “LGInterface (Java Native Code Interface for the Link Grammar API),” [Online]. Available: http://chrisjordan.ca/data/proj/LG-windows.rar, 2007
[Jurafsky 2000] Jurafsky, D. and Martin, J. H., “SPEECH and LANGUAGE PROCESSING: An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition,” Prentice Hall, ISBN: 0-13-095069-6, 2000
[Katz 1993] Katz, B., “START (SynTactic Analysis using Reversible Transformations),” MIT's Artificial Intelligence Laboratory, [Online]. Available:  http://start.csail.mit.edu/, 1993
[Li 2001] Li, J., Zhang, L. and Yu, Y., “Learning to generate semantic annotation for domain specific sentences,” In the Workshop on Knowledge Markup and Semantic Annotation at the 1st International Conference on Knowledge Capture (K-CAP 2001), Victoria, B.C., Canada, 2001 
[Li 2006] Li, X. and Roth, D., “Learning question classifiers: The role of semantic information,” Natural Language Engineering, vol. 12, no. 3, 2006, pp.229-249
[McGuinness 2004] McGuinness, D. L. and Harmelen, F. V., “OWL Web Ontology Language Overview,” W3C Recommendation, [Online]. Available: http://www.w3.org/TR/owl-features/, 2004
[Miller 2003] Miller, G. A., “WordNet, a lexical database for the English language,” Princeton University, Cognitive Science Laboratory, [Online]. Available: http://www.cogsci.princeton.edu/~wn/, 2003
[Sleator 1991] Sleator, D. and Temperley, D., “Parsing English with a Link Grammar,” Carnegie Mellon University Computer Science technical report, CMU-CS-91-196, 1991
[Sleator 2003] Sleator, D., Temperley, D., and Lafferty, J., “Link Grammar,” Carnegie Mellon University, School of Computer Science, [Online]. Available:  http://www.link.cs.cmu.edu/link/, 2003
[Toutanova 2000] Toutanova, K. and Manning, C. D., “Enriching the Knowledge Sources Used in a Maximum Entropy Part-of-Speech Tagger,” in Proceedings of the Joint SIGDAT Conference on Empirical Methods in Natural Language Processing and Very Large Corpora (EMNLP/VLC-2000), Hong Kong, 2000, pp.63-70
[Wang 2005] Wang, Y.-H., Wang, W.-N., and Lin, C.-H., “An Intelligent Semantic Agent for E-Learning Message Communication,” Journal of Information Science and Engineering, Special Issue on E-Learning, vol. 21, no. 5,  pp.1031-1051, 2005
[Wang 2006a] Wang, Y.-H., Huang, C.-C. and Wang, W.-N., “A Semantic-Aware Methodology Adapt to e-Learning Environment,” Journal of Computers, vol. 17, no. 3, pp.41-54, 2006
[Wang 2006b] Wang, Y,-H., Liao, C.-H., Huang, S.-H. and Wang, W.-N, “Application Architecture of Semantic QA System,” Journal of Computers, vol. 17, no. 3, pp.33-39, 2006
[Wang 2007] Wang, Y.-H., Wang, W.-N. and Huang, C.-C., “Enhanced Semantic Question Answering System for e-Learning Environment,” Workshops Proceedings of 21st International Conference on Advanced Information Networking and Applications (AINA 2007), Niagara Falls, Ontario, Canada, May 21 - 23, 2007, vol. 2, pp. 1023-1028 
 [Weibel 1995] Weibel, S., Godby, J. Miller, E. and Daniel, R., “OCLC/NCSA metadata workshop report” Dublin, OH: OCLC Online Computer Library Center. [Online]. Available: http://www.oclc.org:5046/oclc/research/publications/weibel/metadata/dublin_core_report.html, 1995
[Wikipedia 2001] Wikipedia, the free encyclopedia, [Online]. Available:  http://en.wikipedia.org/, 2001
[Yen 2007] Yen, Y.-H., “Semantic-aware Ontology for QA System: a Statistical Approach,” M.S. thesis, Department of Computer Science & Information Engineering, Tamkang University, Taiwan, 2007