DEVELOP A WEB-BASED DATA ACQUISITION TOOL FOR DATA MINING STUDIES OF DRUG THERAPIES

• Main Authors: Wei-Chihh Liu, 劉威志
• Other Authors: Yung-Jin Lee
• Format: Others
• Language: en_US
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/91061939216319144127

碩士 === 高雄醫學大學 === 臨床藥學研究所碩士班 === 95 === This study aimed to develop a web-based data acquisition tool for data mining studies of drug therapies inspired by a 2004 study. To combine two predominant techniques, WWW and data mining are the bases of this study. Computer programming languages such as HTML, PHP, MySQL and JavaScript were used to develop the framework and the interaction between the project managers and this tool. EasyPHP was installed on a workstation PC, and was used to set up the client/server with PHP, MySQL, and Apache servers. A Linux PC with OS Fedora Core 4 was installed and served as a remote server for this tool. The Linux PC also served as a web server which allowed experts to log in to complete a project. Exported Weka ARFF file is the final step of this project, which enables project managers to apply Weka to data mining analysis. We successfully reproduced a 2004 study with this tool, including the questionnaires (simulated and generated) and the on-line expert log-in website (collect experts’ decisions and opinions). Moreover, it may shorten the time taken to construct a project using this tool. It may allow the project managers to define a project on their own demand. The project managers no longer need to manually build a project but can use this integrated and web-based tool to quickly complete a setup for a research project.