應用文字探勘技術於客訴留言品質及分類管理之研究

Use the technique of Text Mining in Customer Complaints’ Quality and Classification Management research

柯秀奎(Shiou-Kuei Ke)

指導教授 ： 羅淑娟

國立臺北科技大學/管理學院/工業工程與管理系所/碩士(2005年)

摘要

本研究希望透過一自動客訴留言分類的機制，將會員的留言藉由文字探勘的技術和分類機制將其自動的分門別類，使客訴能有效率且較準確地配屬於合適的客服人員或資訊工程師，透過客訴留言的自動分類機制來加速處理會員在使用上不懂的疑問或客戶抱怨。而透過此系統不但能提升顧客的滿意度也可降低人工瀏覽客訴進行問題配屬的人力，在網站經營有限的人力上，提高人員的生產力及工作效率。
此外，我們也利用自動客訴留言分類機制，來管理網站服務的品質，透過管制圖概念及文字探勘技術的結合，針對客訴留言內容進行監控，藉由分類後屬於顧客抱怨的留言數量，繪製成管制圖，以用來維持及改善社群網站的品質，達到持續吸引新會員加入及避免舊客戶流失的目的。

關鍵字

文字探勘 ； 客訴留言 ； 文章分類 ； 不良率管制圖 ； 檢驗誤差

並列摘要

This research, wishes that through the program of Customer Complaints Automatic Classification System automatically classify the message that had been left by customers on using the technique of Text mining. And it is able to let the Customer Information Services Centre to directly solve the problems by the customers on system functioning, for it can improve the percentage of satisfaction of the customers. By using this automatic classify/identify system, it is not necessary needed anymore to receive information from the Customer Information Services Centre because the Engineer of the Information Centre and the seller will be able to receive the first hand information directly, and immediately solve out all kind of the Network system problems that may occur and integrate about the demand of the customers and the upgrade of the Network system.
Beside, we also use the Customers Complaints Automatic Classification System to control and maintain the quality of service on the Net, through the concept of control chart cycle, it is focusing on the customers message content, after classified and the amount of message that belong to the customer complaint, will be draw out into control chart graphic and is used on improving and maintaining the quality of the public web-site, and this is also being used for reaching the target of attracting new customers joining in and avoiding of loosing the old customers.

並列關鍵字

Text Mining ； Customer Complaints ； Document Classification ； P Chart ； Inspection Error

參考文獻

[1] 蔡厚灼，客訴文件探勘系統，碩士論文，成功大學資訊管理研究所，台南，2003

[5] 張如瑩，多語系平行關鍵頁搜尋引擎之設計與建構，碩士論文，元智大學資訊管理研究所，桃園，2001

[15] Stevenson William J.，黃崇興譯，作業管理，台北:美商麥格羅。希爾國際股份有限公司 台灣分公司，第七版,2002

[24] H. S. Heaps, “Information retrieval, computational and theoretical aspects,” Academic Press,1978

[25] Meadow and T. Charles , “Text Information Retrieval Systems,” Academic Press,1992