Using Data Mining to Analyze Donors’ Behaviors - The Case of Apple Foundation

碩士 === 國立高雄師範大學 === 事業經營學系 === 106 === In the early stage, Taiwan's nonprofit organizations flourished, and the establishment of private nonprofit organizations in Taiwan was charitable. The main resource of nonprofit organizations came from donation income. Therefore, in order to enhance the p...

Full description

Bibliographic Details
Main Authors: HO,HAO-JUI, 何浩睿
Other Authors: Chen,Long-Hui
Format: Others
Language:zh-TW
Published: 2018
Online Access:http://ndltd.ncl.edu.tw/handle/273myz
Description
Summary:碩士 === 國立高雄師範大學 === 事業經營學系 === 106 === In the early stage, Taiwan's nonprofit organizations flourished, and the establishment of private nonprofit organizations in Taiwan was charitable. The main resource of nonprofit organizations came from donation income. Therefore, in order to enhance the public's recognition of nonprofit organizations, the mission and operation of nonprofit organizations need to be promoted. This study is to explore the case of the Apple Daily Charity Foundation that posted the donor case report in the Apple Daily and the website. The Apple Foundation opens donations’s data in the transparent website which includes all the details of the donation progress report and the donors. Therefore, this study addresses 1,984 donors’ cases of open data from 2010 to 2016. The following three parts are explored through the statistical method and decision tree: (1) To explore the relationship between the relevant variables and contributions received. (2) To explore the relationship between the details of the donor's background variable and the amount of the donation. (3) To use decision tree method to construct case report title rule type that affects the amount of donation. The results of the study are as followed: 1. “Donors classified” and “Donors surname” lead to a significant difference in the donation results. 2. “Date of publication” and “month of publication” lead to a significant difference in the donation results. 3. By using the decision tree in constructing prediction model of donation results, thirteen types of the rules could be found in cases reported title for influence donation results, types of the rules then further are grouped into three hight amount of donation, four medium amount of donation and six low amount of donation groups. According to the group, this stidy could describe the case report of the caregiver, caregiver, disease and other titles information, and suggestions were then proposed and given. In addition, the results found that individual contributions accounted for the largest proportion of the total amount of donations, but the highest amount of each contribution is business contributions. About the month of publication of donor case, the contributions accounted in December are higher than other months, and the donor case title will affect the amount of the donation results. In conclusion, the Non-profit organizations can refer to the practices and conclusions of this study as a basis for their contribution analysis and how to raise a number of donations.