Scholarlink

A key challenge university students face is understanding just what they need to learn from the semester of learning activities. To address this, we introduced students to the idea of maintaining their own open learner model of their learning progress. We provided students with: a light-weight hierarchical Open Learner Model (OLM) ontology; class activities on its use; and the OLiMent chatbot, designed to help students use the OLM ontology. We report a study of 227 students’ use of OLiMent in a third-year subject on human-centred aspects of data analytics. We analysed discretionary use, finding that 51% of the students used it before the final exam, and length of user messages correlated with the exam mark. Qualitative analysis of OliMent transcripts from prescribed activities revealed frequent self-reflection, with genuine attempts at self-assessment in 89% of valid submissions, use of the OLM ontology terms in 96% and reflective behaviour strongly correlated with exam performance. In the mid-semester survey, 70% of students rated OLiMent as useful. This rose to 77% at the end of the semester. Our key contributions are: (1) a new way to enable students to become active Open Learner Modellers, based on reflecting on their learning progress; (2) Design of a lightweight OLM ontology and the OLiMent chatbot to scaffold reflection on learning progress; (3) insights from semester-long OLiMent use and implications for future use of the approach.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

OLiMent: Conversations About Open Learner Modelling to Help Learners Understand and Self-assess Learning Goals

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Annie Yuan,
  • Andrew Fang,
  • Danny Liu,
  • Judy Kay

摘要

A key challenge university students face is understanding just what they need to learn from the semester of learning activities. To address this, we introduced students to the idea of maintaining their own open learner model of their learning progress. We provided students with: a light-weight hierarchical Open Learner Model (OLM) ontology; class activities on its use; and the OLiMent chatbot, designed to help students use the OLM ontology. We report a study of 227 students’ use of OLiMent in a third-year subject on human-centred aspects of data analytics. We analysed discretionary use, finding that 51% of the students used it before the final exam, and length of user messages correlated with the exam mark. Qualitative analysis of OliMent transcripts from prescribed activities revealed frequent self-reflection, with genuine attempts at self-assessment in 89% of valid submissions, use of the OLM ontology terms in 96% and reflective behaviour strongly correlated with exam performance. In the mid-semester survey, 70% of students rated OLiMent as useful. This rose to 77% at the end of the semester. Our key contributions are: (1) a new way to enable students to become active Open Learner Modellers, based on reflecting on their learning progress; (2) Design of a lightweight OLM ontology and the OLiMent chatbot to scaffold reflection on learning progress; (3) insights from semester-long OLiMent use and implications for future use of the approach.