Constructing Educational Concept Maps with Multiple Relationships from Multi-Source Data

Xiaoqing Huang,Qi Liu,Chao Wang,Haoyu Han,Jianhui Ma,Enhong Chen,Yu Su,Shijin Wang

DOI: https://doi.org/10.1109/icdm.2019.00132

2019-01-01

Abstract:Concept map is an useful tool to help people organize and improve knowledge. Particularly in educational domain, it is beneficial for students and teachers to improve the learning and teaching quality. Traditionally, manual educational concept maps, provided by teachers, are quite time-consuming and limited to teachers' experience. Thus, it is meaningful to automatically construct high-quality concept maps. However, existing data-driven solutions only focus on either separate data source or single pedagogic relationship, which are not sufficient to satisfy actual demands. To this end, we propose a novel framework, named Extracting Multiple Relationships Concept Map (EMRCM), to construct multiple relations concept maps from Multi-source Data. Specifically, we design various targeted evidences to explore diverse information of multi-source data from different perspectives. Then, we employ three classic classifiers to bulid the predictive model for extracting key concepts and multiple concept relationships using the proposed evidences. We create a real dataset for empirically studying this problem. Extensive experiments on a real-world dataset show the effectiveness of our method.