由於目前臺灣新移民子女的人數日益增多,然而文獻中發現新移民子女若接受現有的輔導措施越多者,其學業表現就越低。因此,為了加強新移民子女的學習能力及找出一套適當的輔導機制以達因材施教的效果,本研究以新移民子女為研究對象,研究新移民子女對於數學科的學習狀況之研究。在研究中首先將以資料探勘的觀點探討影響新移民子女對於學習數學科意圖的原因,再進而對新移民子女對於數學科學習意圖進行分類,期望未來能在第一時間點就能針對新移民子女的特性,進行差異化的輔導,以提升其數學科學習成效。根據研究結果,本研究不只可找到影響新移民子女數學科學習意圖的原因,更可使用k-最近鄰法可有效找出學習意圖較低的學生。本研究建議未來在找到學習意圖較低並輔導新移民子女數學科學習時,可透過以下三建議將可提高其學習成效(1)營造快樂學習的環境(2)找一位標竿學生讓新移民子女做為學習的對象(3)避免新移民子女在身體不舒服或心理不開心時學習數學。
Recently, the number of new immigrants' children have become more and more in Taiwan. However, according to literatures, new immigrants' children's academic performance is lower when they are taken too much student counseling or existing traditional student counseling in school. In this study, we use data mining technique to find the main effect of learning intention in Mathematics for new immigrants' children and build the classification model for learning intention in Mathematics of new immigrants' children. According to the results, this study can search the main effect of learning intention and find the low learning intention in Mathematics for new immigrants' children by k-nearest neighbors. Finally, we suggest when we find the low learning intention's children, we can give a suitable student counseling for new immigrants' children in the future.