學習者個人與學習特質相關性分析

The correlation analysis of learners' characteristics

邱上勻(Shang-Yun Chiu)

指導教授 ： 林育慈

國立暨南國際大學/科技學院/資訊工程學系/碩士(2011年)

摘要

資訊與網路科技的發展使得線上學習日益普及，在無法面授教學的線上學習環境中，適性化的功能尤其重要，根據不同學習者的特質，提供適合於該學習者的學習功能與教材，以達到有效學習之目的。然而，一般的線上學習平台只擷取學習者簡單的個人資料，缺乏學習者學習特質相關資訊，無法提供適性化學習系統足夠的個別化訊息。為了達到根據學習者個別化學習特質提供適性化學習環境的目的，本研究針對學習者個人與學習特質之相關性分析，以期透過簡單個人特質便能獲知學習特質，進一步提供以學習特質進行適性化教學應用之所需。研究中分析之個人特質包含血型、星座、與性別，學習特質則包含學習評量結果、評量作答、與學習風格。
對於純量之學習者特質資料，我們可以傳統之相關分析方法將兩種特質進行分析，然而向量或非順序性資料則無法以傳統方式為之。為了對向量與非順序性資料進行相關性分析，本研究提出相關性分析演算法，先將擷取自學習者特質之特徵向量利用自組映射圖網路分組，再根據幾種分析機制：基於直方圖之相關性分析、基於特徵比例之相關性分析、與主軸分析，探討未分組與組內之資料特性，以了解兩個不同特質之相關性。我們並將各種學生特質的相關性視覺化以提供教師教學改進之用。

關鍵字

學生特質 ； 個人特質 ； 學習特質 ； 相關性分析 ； 餘弦相似性

並列摘要

With the rapid development of information and network technology, web-based learning becomes a new trend of learning. In the on-line learning environment, adaptability is an important issue for the purpose of providing suitable functions or content for different learners for effective learning. However, most on-line learning systems collect only general personal attributes from the input form, but not include learning characteristics, which cannot provide sufficient information for adaptive learning. In order to mine the learning characteristics of students from general user information, this study provide a correlation analysis of various characteristics of learners for understanding the relationships between different user information, so that useful learning characteristics can be found through general personal attributes. The learner characteristics considered in this work include blood types, constellations, sex, assessment results, assessment answering types, and learning styles.
For sequential scalar data, many existing correlation metrics can be applied to analysis them. However, some data contain non-sequential elements or vectors. Therefore, we propose a correlation analysis mechanism to understand relationships between non-sequential or vector data. The features are classified by the Self Organizing Map at first, and then the correlations are estimated by discussing the between-group and within-group distribution based on histogram-based analysis, feature ratio analysis, and principal component analysis. At last, the produced correlation analysis results are visualized to provide teachers with a correlation map for improving their teaching.

並列關鍵字

Learner characteristics ； Personal qualities ； Correlation analysis ； Cosine similarity

參考文獻

[1] Dodridge, M. (1999). Learning Outcomes and Their Assessment in Higher Education. Engineering Science and Education Journal, 8(4), 161-168.

[2] Marshall, P. (1994). Open Exams an Experiment in Student Assessment. Engineering Science and Education Journal, 3(1), 15-20.

[3] Antao, B.A.A., Brodersen, A.J., Bourne, J.R., & Cantwell, J.R. (1992). Building Intelligent Tutorial Systems for Teaching Simulation in Engineering Education. IEEE Trans. Educ., 35(1), 50-56.

[4] Smaill, C.R. (2005). The Implementation and Evaluation of OASIS: A Web-Based Learning and Assessment Tool for Large Classes. IEEE Trans. Educ., 48(4), 658-663.

[5] Guzman, E., & Conejo, R. (2005). Self-Assessment in A Feasible, Adaptive Web-Based Testing System. IEEE Trans. Educ., 48(4), 688-695.