A Study of Customer Value Analysis on LCD Panel After-Sale Service by Data Mining Method

碩士 === 國立成功大學 === 工學院工程管理碩士在職專班 === 100 === The TFT-LCD (Thin Film Transistor-Liquid Crystal Display) industry of Taiwan has been owned an important role in the worldwide international market. Nevertheless, its customer relationship management is still applying the practical experience approach with...

Full description

Bibliographic Details
Main Authors: Yu-TzuTsai, 蔡玉慈
Other Authors: Tse-Sheng Chen
Format: Others
Language:zh-TW
Published: 2012
Online Access:http://ndltd.ncl.edu.tw/handle/54939552208009345840
Description
Summary:碩士 === 國立成功大學 === 工學院工程管理碩士在職專班 === 100 === The TFT-LCD (Thin Film Transistor-Liquid Crystal Display) industry of Taiwan has been owned an important role in the worldwide international market. Nevertheless, its customer relationship management is still applying the practical experience approach with no maturity methods. The technical services of the TFT-LCD industry belong in a tightly mutual interaction model with the customers. However, in the era of the low product profit, a higher customer interaction industry with the breakthrough mass production technology, its price will be decreased gradually year by year. Therefore, a study of customer value analysis on TFT-LCD panel after-sale service is quite importance. Thus, the satisfied needs of the customers will be fulfilled and their difficult problems can be solved. After-sale service is the last item in the customer relationship management. The proposed OLAP method has been studied to find out the major factors affecting customer value of the TFT-LCD industry by running the decision tree analysis. We applied data mining technique to understand and forecast quickly for the customer value items. Thus, it can be changed and viewed from the cost unit rather than profit one. We studied and collected the relative items from TFT-LCD of after-sale service data firstly. We preprocessed and transferred them into data warehouse to fit the computation of the data mining technique secondly. Meanwhile, we also matched up the original Manufacturing Executive System database to build the C4.5 decision tree and running by WEKA software package. Some rules of the decision tree are discovered for affecting customer value of the TFT-LCD industry. It provided a reference model for making an appropriate decision for every product and customer within shorten time including market profitability, quality status and maintenance cost. Finally, the study will provide some valuable ideas for the TFT-LCD industry in forecasting and monitoring strategies as well as investment decision making.