臺灣博碩士論文加值系統

由於網際網路的便利性，因此發展出各式各樣的電子商務活動及內容提供網站(ICP)。使用者只要透過搜尋引擎就可以找到許多相關的商品、服務或內容資訊。然而對於搜尋引擎所查詢到的資料廣度及深度都巳經有非常不錯的結果，但是面對網頁內容不一，而且又不是以標準的格式來呈現，也造成資訊蒐集及整合工作上會有相當大的負擔。
因此本研究希望運用代理人技術，並透過學習式機制，能夠從文件之中擷取出有用的部份，以創造出資訊本身的價值。其中主要結合智慧搜尋、資訊過濾、資訊擷取、規則學習的方法，以提升資訊擷取的品質，達到資訊蒐集目的。最後根據研究的內容，建置成購物代理人系統，並以網路上的購物資訊作為資料來源，讓使用者能多樣化的查詢特定商品資訊並提供個人化資訊整合服務。同時以評估系統對於資料的收集能力和擷取的正確率，來作為驗證本研究可行性和實用性之所在。

In recent years, due to the emerging technologies on the Internet, various Internet services are available including e-commerce activities and Internet content providers. It is convenient to find the desired product information from the Internet by utilizing various search engines available. However, there exist no standards for the content providers in presenting their valuable data in a structured way. Therefore, to effectively extract specific information and collect data from the semi-structured Internet documents has become a problem.
In this study, a shopping agent is designed and implemented. It is an intelligent agent that does information extraction and has a global supervised-learning mechanism. In other words, this agent learns every time it provides its service to a user. The knowledge is accumulated during the runtime of the system. Whatever it learned from one user is reusable globally by others that request the similar services from the same agent. This is a very important concept here since it learns once and works forever unless the reference template is modified. Various experiments are conducted which tried to simulate searching for some merchandise. The experimental results are promising and the system has demonstrated its practicability as a great shopping assistant.

書名頁i
論文口試委員審定書ii
授權書iii
中文摘要vi
英文摘要vii
誌謝viii
目錄ix
表目錄xii
圖目錄xiii
一、緒論1
1.1.研究背景1
1.2.研究動機1
1.3.研究目的3
1.4.研究範圍4
二、文獻探討5
2.1.資訊檢索與擷取5
2.2.文件相關度比對6
2.3.網頁資訊擷取8
2.4.資訊整合包覆程式10
2.5.網際網路購物代理人系統10
三、研究內容與方法11
3.1.研究架構11
3.2.智慧搜尋代理人12
3.3.資訊過濾處理14
3.3.1.權重學習訓練及更新法則15
3.3.2.文件相似度比對17
3.4.資訊擷取處理19
3.4.1.資料分析擷取21
3.4.2.文件結構擷取23
3.4.3.規則學習擷取25
3.5.個人化資訊整合33
四、系統實作與測試35
4.1.系統架構35
4.2.系統開發環境36
4.3.系統實作37
4.3.1.使用者查詢介面37
4.3.2.線上規則學習40
4.3.3.工作控制模組41
4.3.4.資訊搜尋代理人(Query Agent)42
4.3.5.資訊過濾代理人(Filter Agent)43
4.3.6.資訊收集代理人(Gather Agent)44
4.3.7.類神經網路文件特徵權重學習46
4.4.系統測試48
4.4.1.測試評估方法48
4.4.2.測試結果49
4.4.3.結果分析54
五、結論57
5.1.研究成果57
5.2.研究限制59
5.3.未來工作61
參考文獻62
附錄一65

1.Kalakota, R., and Whinston, A. B., Frontiers of Electronic Commerce, Reading, Mass., Addision-Wesley, 1996.
2.Kambil, A., "Doing Business in the Wired World," IEEE Computer, Vol.3, No.5, pp.56-61, May 1997.
3.Cowie, J. and.Lehnert W., "Information Extraction", Communication of ACM, Vol 39, No1, pp.80-91, January 1996.
4.Kushmerik, N, "Gleaning the Web", IEEE Intelligent Systems, Volume: 14 2 , Page(s): 20 -22, 1999.
5.San, M., "Intelligent agents on the Internet and Web", TENCON ''98. IEEE Region 10 International Conference on Global Connectivity in Engery, Computer, Communication and Control, Vol.1, pp.97-102.
6.Faloutsos, C. and Oard D., "A survey of Information Retrieval and Filtering Methods", University of Maryland College Park, CS-TR-3514, 1995. Also available at http://www.enee.umd.edu/medlab/filter/papers/survey.ps
7.Nie J., Briscbois M., and X. Ren, "Template-Based Information Mining from HTML." In: Proceedings of AAAI-97, Providence, USA, 1996.
8.莊水龍，「助理型軟體資訊擷取技術：樹狀結構 HTML 文件樣板之自動產生法」，國立臺灣大學資訊工程學研究所，1998。
9.Chun-Nan Hsu, "Initial results on wrapping semistructured web pages with finite-state transducers and contextual rules", In Proceedings of AAAI-98 Workshop on AI and Information Integration, Technical Report WS-98-01, Menlo Park, CA, 1998.
10.Yanlei Diao, Hongjun Lu, Songting Chen, Zengping Tain, "Toward Learning Based Web Query Processing", VLDB 2000, 317-328, 2000.
11.Wolfgang May, "An integrated Architecture for Exploring, Wrapping, Mediating and Restructuring Information from the Web", Database Conference, 2000. ADC 2000. Proceedings. 11th Australasian, Page(s): 82 -89, 2000.
12.Ling Liu, Calton Pu, Wei Han, "XWRAP an XML-enabled Wapper Construction System for Web Information Sources", Data Engineering, 2000. Proceedings. 16th International Conference, Page(s): 611 -621, 2000.
13.陳光華，電子文件自動處理之研究，學生書局，台北，1999。
14.曾元顯，"關鍵詞自動擷取技術之探討"，中國圖書館學會會訊，5 卷 3 期（106），1997。
15.Salton and McGill, Introduction to Modern Information Retrieval, McGraw Hill Book Co, 1989.
16.Lee H. C, Dagli C. H., Ercal F., and Ozbayoglu A. M., "SimNet: A Parallel Neuro-Fuzzy Paradigm for Data Clustering", OAI Neural Networks Symposium and Workshop(OAINN ''95), Athens, Ohio, USA, 1995.
17.許加文、李錫捷，「網際網路搜尋過濾系統[一個「關鍵頁」超搜尋智慧型代理引擎]，第九屆國際資訊管理學術研討會，中壢，台灣，1998。
18.謝超煒，「網際網路資訊擷取過濾系統-中文關鍵頁超搜尋代理人」，私立元智大學資訊管理研究所，碩士論文，2000。
19.S. Aggarwal, F. Hung, Weiyi Meng, "WIRE-a WWW-based information retrieval and extraction system", IEEE Database and Expert Systems Applications, Page(s): 887 -892, 1998.
20.黃宏軒，「全球資訊網資訊整合可程式化包覆程式的設計與實作」，國立臺灣大學資訊工程學研究所，碩士論文，1999。
21.陳昱彰，「個人化網際網路電子購物代理人」，國立交通大學資訊工程學系，碩士論文，1997。
22.簡逸音，「個人化技術在網際網路購物代理人之研究」，中華大學電機工程學系碩士班，碩士論文，1999。
23.顏逸品，「網際網路半結構化資料之蒐集與整合研究」，國立中央大學資訊管理研究所，碩士論文，1999。
24.PCHome Shopping Guide，http://shoppingguide.ithome.com.tw/pricecompo/socket7mainboard_index.html
25.Dietrich H. Schuschel and Chun-Nan Hsu, "A weight analysis-based wrapper approach to neural nets feature selection", In Proceedings of the 10th IEEE International Conference on Tools with AI, Taipei, Taiwan, 1998.
26.Chun-Nan Hsu and Ming-Tzung Dung, "Generating finite-state transducers for semistructured data extraction from the web", Journal of Information Systems, Special Issue on Semistructured Data, Volume 23, Number 8, 1998.
27.YAHOO購物網站，http://advanced.search.shopping.yahoo.com/shop/cpda?manufacturer=Palm&x=1
28.楊雪花，國語常用虛詞及其台語對應詞釋例，1997
29.Robertson E. S., "The Parametric Description of Retrieval Tests", Journal of Documentation 25:1, pp.1-27.
30.XTM TopicMaps.org，http://www.topicmaps.org/