Statistical Analysis of the Data from University Assembly Meetings and the Opinion Web System of the National Taiwan University

• Main Authors: Li-Wei Yao, 姚力維
• Other Authors: 劉仁沛
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/11915268728746744236

碩士 === 國立臺灣大學 === 農藝學研究所 === 102 === 　　Technology research firm International Data Corporation estimates that the growth rate of the annual amount of global data is 50%. Data has doubled every two years. It is very important to use the techniques and methods of extracting meaningful information from data. There are many other kinds of data in the world such as audio, video, PDF, and web pages, in addition to text or data stored in the database. 　　This study is focused on the data from web pages on internet: University Assembly Meetings from National Taiwan University between 2001 and 2012 and the Opinion Web System of the National Taiwan University between 2005 and 2012. The methodology involves collection, transcription, and quantitation of the web data which were analyzed by statistical methods such as descriptive statistics, graphic presentation, analysis of variance, multiple comparisons, chi-square test, and nonparametric procedure. Our study objectives are to compare processing duration of proposed bills sponsored by different categories of sponsors for University Assembly Meetings and to compare the processing duration and popularity of opinions by different categories of opinion areas and administrative units for the Opinion Web System of the National Taiwan University. 　　The results show that the statistical significant differences were found in proposed bills and processing efficiency among administrative units, faculty and students in University Assembly Meetings from National Taiwan University. The mean processing time of the proposed bills sponsored of faculty and students is statistically longer than that sponsored by administrative units. It seriously affects the interest of faculty and students. For the opinion web system of the National Taiwan University, we compared the processing duration and popularity between categories of opinions, subjects of opinions, and identification classes of the proposed of the opinions. Except for the winter and summer vacations, an increasing trend is observed in opinion volume.