標題: 利用項目反應理論推薦個人化課程
Personalized Course Recommendation using Item Response Theory
作者: 石昕秀
盧鴻興
SHIH, HSIN-HSIU
Lu, Horng-Shing
統計學研究所
關鍵字: 項目反應理論;非線性混合模型;個人化學習;Item response theory;Nonlinear mixed model;Personalized learning
公開日期: 2017
摘要: 個人化學習在教育領域是一個很重要的研究議題。目前,很多學習系統認為應考慮學習者的能力作為個人化學習的重要依據。有些個人化學習系統依據項目反應理論;項目反應理論用於描述項目困難度與學習者能力之間的關係。然而,這些個人化學習系統 是使用基礎的項目反應模型,並未考量項目本身的特徵。本研究使用較複雜的項目反應模型,稱為非線性混合模型,來進行課程推薦。非線性混合模型可考量項目本身的特徵。結果顯示,非線性混合模型可以提升模型準確度,推薦較符合學習者能力的課程,並避免推薦過於困難的課程造成學習者過多的負擔。
Personalized learning is an important research topic in education. Many learning systems consider learner’s ability a significant factor in providing personalized learning services. Oftentimes, these systems employ item response theory to describe the relationship between item difficulty and learner’s ability. However, most learning systems are based on basic item response models which don’t consider in the item features. This study proposes a personalized course recommendation based on the extended item response model, one of the nonlinear mixed models, which takes into account the item features. Results show that the nonlinear mixed model improved the model accuracy and provides better recommendation.
URI: http://etd.lib.nctu.edu.tw/cdrfb3/record/nctu/#GT070452609
http://hdl.handle.net/11536/140932
Appears in Collections:Thesis