电子学习数据分析

电子学习数据分析

  • 电子学习数据分析
    ART/190CP
    20150127 - 20160726
    12921

    陈国娴博士
    There are five major deliverables in this project 1. Learning Data Collection i. Define an open interface to be connected to third party LMS and publishers’ server ii. Information Visualization software tool a) presentation of data in the form of learning dashboards which provide overview learning data through data visualization tools b) Teachers may be able to select the types of data to be presented 2. Quiz-based Learning Data Analysis Engine i. Models definition based on EDB Learning data a) Models should be able to identify the areas of weakness for the students b) Models constructed based on more than 5000 training questions (including questions data and student profile data) ii. Learning analysis engine implemented in software and applied in primary school mathematics subject evaluation iii. Automatic feedback capability - new set of questions targeting the weakness can be generated automatically a) Identify areas of weakness of students, and to provide students with tailored learning pathways, or assessment materials (90% of the identifications is in agreement with school teachers) b) 90% of the auto generated set of questions are accepted by school teachers, which will be validated in the school trial. 3. Context based Text Mining/Text Analysis Software Engine This focuses mainly on primary school students' Chinese writing i. Develop text mining algorithm specific for identifying Colloquial Cantonese (software) ii. The algorithm will be able to identify 80% of the common errors in the writings of primary school students 4. Learning Analytics Server i. Connection to third party LMS, publisher sites (at least two sites) based on the open interface ii. With text mining/text analysis engine and learning data analysis engines built in iii. Provide instant suggestions on the correction on common errors with accuracy of 80% 5. EDB School trial of learning analytics server i. With at least 5 schools ii. Report generation
    刘文健博士 吴志刚先生 王雁晨女士 洪丹博士 芦运照博士 陈文财先生 何嘉颖女士 莊秋萍女士 杨逸先生 王蕊小姐 陈宜廉先生 林爱华女士
    教育局

    在云端学习环境中,不同的网站提供了形形色色的学习活动。这些活动数据,对分析学生的学习行为的全面性尤其重要。因此,有需要去定义一个标準的接口,令到网站之间的数据交换更简单和快捷。 收集数据後,可作进一步的分析。根据活动的性质,两种类型的分析可以进行。首先,以测验为基础的分析。它采用数据挖掘和分析技术来评估学生在练习或测验中的表现,去识别个别学生的弱点。一组新的问题或补充材料的建议,针对著学生的弱点,可以自动产生。这将大大提高个人化的学习经验。 第二,基於文章内容的分析。它利用文本挖掘和分析技术,针对学生在论坛或文章的写作,辨别口语化的中文用词,即时指出学生的错误并作出修改建议,不需教师太多的干预。