| Author: | Chong, Kin-tak David |
| Title: | To improve service center operations by using data mining technology |
| Degree: | M.Sc. |
| Year: | 2007 |
| Subject: | Hong Kong Polytechnic University -- Dissertations. Customer services -- Management -- Data processing. Data mining. |
| Department: | Department of Computing |
| Pages: | 94 leaves : ill. ; 30 cm. |
| Language: | English |
| Abstract: | The purpose of this study is to use data mining techniques to improve part of the operations in a machinery trading company's service center. The two areas of operation to be studied include: (1) To automate the assignment of service orders to engineers. (2) To provide manpower requirement forecasting for each coming month. First, the current models of operation will be described. Problems of working under the current models will be identified. Then different types of data mining modeling method will be studied and the most appropriate technique for this project will be selected. After the model building, testing and validation process, new models of operations to achieve the original goals will be proposed. The methodology used in this study is called CRISP-DM, a widely adopted method for data mining projects. This methodology includes six steps - Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation and Deployment. This methodology is clear and easy to follow and suitable for large or small projects. After this study, a new model for service orders assignment is built. Based on this model, Customer Service Officer (CSO) can assign jobs to engineers with minimum, if any, assistance from engineer supervisors. Basically, a new data mining model will be built based on the service history of a service group. The model is a classification model built by the CHAID algorithm. Customer Service Officers (CSO) only needs to take tomorrow's service order as input and the system will proposed a list of engineers for the orders. Of course, supervisors still have the chance to amend the arrangement, for whatever reasons, if they really want to do so. This system can reduce the reliance on engineer supervisors' skill to assign orders. Another model resulted from this study is for manpower forecasting. This model takes the same input from the Service System for model training. By combining the result from this model (based on past history) and other information, such as engineer list for each group and the holiday request records, the system will inform the users whether the forecast demand would exceed the available manpower. These two applications of data mining tools in service center emphasize one major target - Automation. Automation allows more flexibility to the Company and increase efficiency because there is less reliance on individual's ability or knowledge. The experience in this study shows that data mining is a very powerful tool for business in the area of finding useful information, even automatically. It also shows that there are limitations when using data mining models. For example, one model suit for a certain business segment or user groups may not be useful for others. The model will be outdated due to change of trends. So models must be updated at appropriate interval. Also, the selection of data (in terms of period or data fields) for training the model will be critical for the accuracy / usefulness of the model. Data mining is an iterative process. That means the process might be need to repeat several times before you can have satisfactory results. Also, the current phase may lead you to aware the problems or inadequacy in the previous phase, so you have to start from the previous phase again. As a conclusion, Data Mining is very useful for business improvement. However, this technique must be used very carefully and appropriately in order to obtain useful information and achieve business goals. |
| Rights: | All rights reserved |
| Access: | restricted access |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| b21459824.pdf | For All Users (off-campus access for PolyU Staff & Students only) | 6.62 MB | Adobe PDF | View/Open |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/2467