| Summary: | 碩士 === 國立臺北科技大學 === 工業工程與管理研究所 === 98 === Nowadays, the competitions in society industry are getting fiercer than ever. How to stand out from the crowd in a competitive market to be the leader is the goal in many enterprises. In order to achieve this goal, the companies not only improve the technology, quality, design and cost, but also introduced “customer primarily” to enhance the competitiveness of enterprises; thus many high-tech companies in Taiwan define themselves are “customer-oriented” service industry to achieve world-class enterprise standard.
How to handle customer complaints and questions in a timely and proper manner is an issue of paramount importance in “customer-oriented” businesses, which has accordingly triggered extensive studies in recent years. An online customer service system capable of automatically classifying incoming questions and providing immediate responses will thus be a valuable asset in enhancing service quality and customer satisfaction. To save internal labor costs and time costs, it is necessary to reduce redundant duplication information dealing with the same or frequency questions for the sake of facilitating real-time responses to customers.
The study is proposing the set of semi-automated customer service system with the concept of two-stage approach to carry out: (1). Automatic Classification ︰ First, using text mining techniques to classify the initial customer messages automatically. Second, using the system sends classified messages to the responsible departments respectively.
(2). Customer Problems Automatic Reply︰Application case based reasoning technology, and the case keywords for similarity matching. Compared to customer questions, the highly similar case can immediately reply to customer. Reference to the past solution processing, real-time and reply promptly to customer, will enhance customer satisfaction and corporate loyalty.
|
|---|
