客戶關係管理(Customer Relationship Management, CRM)是近年來相當熱門的話題，主要目的在於改善公司與客戶之間的關係。如何透過「客戶關係管理」來瞭解客戶的需求，提供良好的服務來拓展及保留客戶，為公司創造最大的利益，是目前電信業相當重視的一個問題。為在電信業GSM領域行動電話障礙排除客戶服務的相關研究，卻顯示下列問題：(1)客戶服務資料庫中的GSM領域詞彙知識很不完整且無系統性的管理；(2)客服人員在遇到無法排解之障礙情形時，會利用「其他」值來記錄，使得資料庫存在大量無意義的資料；(3)不同客服人員，對於同一障礙問題，往往有不一致的解答出現；(4)客戶服務資料庫為客服人員人工記錄，難免會有記錄闕漏的情形。
為改善上述問題，我們根據【Liao04】中「資料前置處理器」的概念為基礎，作適當的改進，建構一障礙排除系統，透過此系統，來自動替使用者所遇到的障礙情形尋求解答。在本研究中，我們引進下列技術來幫助解決上述問題：(1)引進知識本體技術來明確規範GSM領域的基本知識及詞彙；(2)引進全面性的資料庫重整技術來將客戶服務資料庫重整為更具意義的資料；(3)引進改良式的MMS關連式規則來進行規則挖掘，解決現實生活資料庫的缺失，並且(4)透過規則驗證技術來確保規則庫的精確及完善。
實驗顯示：(1)全面性的資料庫重整技術能將資料庫整理得更具意義，有利於資料挖掘以及增強規則庫的涵蓋範圍；(2)改良MMS關連式規則，使其更能靈活解決「rare item problem」；(3)新的規則驗證技術不僅能夠確保規則順利鏈接，也能確保規則庫所能涵蓋之障礙情形範圍，更有利於障礙排除之推理工作。

Customer Relationship Management (CRM) is a very hot topic in recent years. The main purpose lies in improving the relationships between companies and customers. How to find out a customer's demand and accordingly offer a good service to recruit a new customer or keep an existent customer through CRM is what the telecommunication industry pays much attention to. A preliminary study on customer service about trouble shooting of GSM cellular systems, however, shows some problems: (1) The knowledge of vocabulary used to describe trouble shooting of GSM cellular systems in the customer servicing database is unorganized and incomplete. (2) A CSR (Customer Service Representative) is tempted to describe an unclear problem with the term “other” when improper vocabulary terms can be used. It renders the customer servicing database filled with a large number of meaningless data. (3) Different CSRs tend to provide different suggestions for the same user-reported problems. (4) The customer servicing database is manually recorded; it is easy to overlook significant data.
Based on the concept of database preprocessing described in [Liao04], this thesis proposes to automatically construct a trouble shooting system to solve the above-mentioned problems. We introduce the following techniques: (1) Ontology technique: we clearly define and organize significant fundamental concepts about GSM in ontology; (2) Comprehensive database refurbishing technique: we refurbish the customer servicing database by replacing both improper description and “other” values with proper terms; (3) Modified-MMS association rule technique: we base on MMS to mine trouble shooting rules to cope with the rare item problem in real life databases; (4) Rule base verification technique: we verify the rules so that rule integrity can be sustained during inference.
Our experiments showed the comprehensive database refurbishing technique did make database more meaningful so that subsequent data mining not only produced more, useful rules but also expanded the coverage of the rule base. The modified MMS association rule mining technique did make it more flexible to solve the rare item problem. Finally, the rule base verification technique not only guaranteed the success of rule chaining, but also increased rule coverage, a result more beneficial to the inference of any trouble shooting system.

【Agra94】 R. Agrawal and R. Srikant, “Fast Algorithms for Mining Association Rules,” Proceedings of the 20th International Conference on Very Large Databases, pp. 487-499, 1994.
【Chie99】 L. F. Chien, “PAT-Tree-Based Adaptive Keyphrase Extraction for Intelligent Chinese Information Retrieval,” Information Processing and Management, Vol. 35, No. 4, pp. 501-521, 1999.
【Dunh03】 M. H. Dunham, “Data Mining Introduction and Advanced Topics, “ Prentice Hall, 2003.
【Dvrr99】 J. Dvrre, P. Gerstl, and R. Seiffert, “Text mining: finding nuggets in mountains of textual data,” Proceedings of the Fifth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, August 1999.
【Grub93】 T.R.Gruber, “A Translation Approach to Portable Ontology Specifications,” Journal of Knowledge Acquistion, Vol.5, pp. 199-220,1993
【Han99】 J. Han, J. Pei, and Y. Yin, “Mining frequent patterns without candidate generation,“ Proceedings of ACM-SIGMOD International Conference on Management of Data, 1999
【Joac96】 T. Joachims, “A Probabilistic Analysis of the Rocchio Algrithm with TFIDF for Text Careagorization,” Technical Report of CMU-CS-96-118, Department of Computer Science, Carnegie Mellon University, Pennsylvania, USA, March 1996.
【Liao04】 I. K. Liao, “Ontology-Supported Data Preprocessing and Mining Techniques for Trouble Shooting of GSM Systems,” Master Thesis, Department of Computer Science and Information Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan, R.O.C., 2004.
【Liu99】 B. Liu , W. Hsu , and Y. Ma, “Mining Association Rules with Multiple Minimum Supports,” Proceedings of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining, p.337-341, August 15-18, 1999.
【Lui00】 C.L. Lui, F.L. Chung, “Discovery of Generalized Association Rules with Multiple Minimum Supports,” Proceedings of the 4th European Conference on Principles of Data Mining and Knowledge Discovery, Vol. 1, pp. 510-515, 2000.
【Mann98】 H. Mannila, “Database methods for data mining.” KDD-98 tutorial, 1998.
【Neto00】 J. L. Neto, A. Santos, C. Kaestner, and A. Freitas, “Document Clustering and Text Summarization,” Proceedings of the Fourth International Conference Practical Applications of Knowledge Discovery and Data Mining (PADD-2000), pp. 41-55, London, January 2000.
【Noy01】 N. F. Noy and D. L. McGuinness, “Ontology Development 101: A Guide to Creating Your First Ontology,” Stanford Knowledge Systems Laboratory Technical Report KSL-01-05 and Stanford Medical Informatics Technical Report SMI-2001-0880, March 2001, Available athttp://www.ksl.stanford.edu/people/dlm/papers/ontology-tutorial-noy-mcguinness.pdf
【Ozel01】 S. Ayse zel and H. Altay Gvenir, “An Algorithm for Mining Association Rules Using Perfect Hashing and Database Pruning,” Proceedings of the Tenth Turkish Symposium on Artificial Intelligence and Neural Networks (TAINN'2001), A. Acan, I. Aybay, and M. Salamah (Eds.), Gazimagusa, T.R.N.C. (June 2001), 257-264.
【Park97】 J. S. Park, M. S. Chen and P. S. Yu, “Using a Hash-Based Method with Transaction Trimming for Mining Association Rules,” IEEE Transactions on Knowledge and Data Engineering, Vol. 9, No. 5, (Sept./Oct. 1997).
【Tsai00】 C. H. Tsai “MMSEG: A World Identification System for Mandarin Chinese Text Based on Two Variants of the Maximum Matching Algorithm,” available at http://www.geocities.com/hao510/mmseg/ , 2000.
【Tsen01】 M.C. Tseng , W.Y. Lin, “Mining Generalized Association Rules with Multiple Minimum Supports,” Proceedings of the Third International Conference on Data Warehousing and Knowledge Discovery, p.11-20, September 05-07, 2001
【Zaki00】 M.J. Zaki, “Generating non-redundant association rules,” Proceedings of the Sixth ACM-SIGKDD International Conference on Knowledge Discovery and Data Mining, New York, NY: ACM, 34-43. 2000.
【Zhen01】 Z. Zheng, R. Kohavi and L. Mason, "Real world performance of association rule algorithm," In Proc. 7-th Int. Conf. on Knowledge Discovery and Data Mining, 2001.
【Zou02】 Q. Zou, W. Chu, J. David, and H. Chiu, “A Pattern Decomposition Algorithm for Data Mining of Frequent Patterns,” Knowledge and Information Systems, Vol. 4, pp. 466-482, 2002.