Abstract
There exist many approaches that enable learning analytics within the educational framework, but their adoption is hindered by the capability of integrating teachers’ inquiries into analysis and by the lack of flexibility to adapt analytics to meet teacher specific needs. This study investigates how to use existing tools to build a framework able to respond teachers’ inquiries when they are posed in natural language, offering the answers from a data-driven perspective. As a result, this paper proposes a framework to integrate teachers’ inquiries into the analysis of a massive dataset using the IBM Watson Analytics tool and presents a case study that validates that framework. The massive dataset contains over 10 years’ student behavior data stored in a LMS of a Chinese university in SQL format with a size of over 1 Giga bytes. The results show that analytics can help to answer generic questions that can be formulated using natural language. We also analyze the impact of the data curation process on the quality of the obtained answers.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of Learning Analytics Adoption at the Classroom Level |
| Publication status | Published - 5 Mar 2018 |
| Event | International Workshop on Orchestrating Learning Analytics (OrLA): Learning Analytics Adoption at the Classroom Level In conjunction with LAK 2018 at The University of Sydney, Australia - University of Sydney, Sydney, Australia Duration: 5 Mar 2018 → 8 Oct 2028 https://sites.google.com/view/orla-ws-2018/home?authuser=0 |
Conference
| Conference | International Workshop on Orchestrating Learning Analytics (OrLA): Learning Analytics Adoption at the Classroom Level In conjunction with LAK 2018 at The University of Sydney, Australia |
|---|---|
| Country/Territory | Australia |
| City | Sydney |
| Period | 5/03/18 → 8/10/28 |
| Internet address |