-
A Transparency Index Framework for AI in Education
Authors:
Muhammad Ali Chaudhry,
Mutlu Cukurova,
Rose Luckin
Abstract:
Numerous AI ethics checklists and frameworks have been proposed focusing on different dimensions of ethical AI such as fairness, explainability, and safety. Yet, no such work has been done on develo** transparent AI systems for real-world educational scenarios. This paper presents a Transparency Index framework that has been iteratively co-designed with different stakeholders of AI in education,…
▽ More
Numerous AI ethics checklists and frameworks have been proposed focusing on different dimensions of ethical AI such as fairness, explainability, and safety. Yet, no such work has been done on develo** transparent AI systems for real-world educational scenarios. This paper presents a Transparency Index framework that has been iteratively co-designed with different stakeholders of AI in education, including educators, ed-tech experts, and AI practitioners. We map the requirements of transparency for different categories of stakeholders of AI in education and demonstrate that transparency considerations are embedded in the entire AI development process from the data collection stage until the AI system is deployed in the real world and iteratively improved. We also demonstrate how transparency enables the implementation of other ethical AI dimensions in Education like interpretability, accountability, and safety. In conclusion, we discuss the directions for future research in this newly emerging field. The main contribution of this study is that it highlights the importance of transparency in develo** AI-powered educational technologies and proposes an index framework for its conceptualization for AI in education.
△ Less
Submitted 9 May, 2022;
originally announced June 2022.
-
Multi-Task Learning based Online Dialogic Instruction Detection with Pre-trained Language Models
Authors:
Yang Hao,
Hang Li,
Wenbiao Ding,
Zhongqin Wu,
Jiliang Tang,
Rose Luckin,
Zitao Liu
Abstract:
In this work, we study computational approaches to detect online dialogic instructions, which are widely used to help students understand learning materials, and build effective study habits. This task is rather challenging due to the widely-varying quality and pedagogical styles of dialogic instructions. To address these challenges, we utilize pre-trained language models, and propose a multi-task…
▽ More
In this work, we study computational approaches to detect online dialogic instructions, which are widely used to help students understand learning materials, and build effective study habits. This task is rather challenging due to the widely-varying quality and pedagogical styles of dialogic instructions. To address these challenges, we utilize pre-trained language models, and propose a multi-task paradigm which enhances the ability to distinguish instances of different classes by enlarging the margin between categories via contrastive loss. Furthermore, we design a strategy to fully exploit the misclassified examples during the training stage. Extensive experiments on a real-world online educational data set demonstrate that our approach achieves superior performance compared to representative baselines. To encourage reproducible results, we make our implementation online available at \url{https://github.com/AIED2021/multitask-dialogic-instruction}.
△ Less
Submitted 15 July, 2021;
originally announced July 2021.
-
A systematic review of physical-digital play technology and developmentally relevant child behaviour
Authors:
Pablo E. Torres,
Philip I. N. Ulrich,
Veronica Cucuiat,
Mutlu Cukurova,
Maria Fercovic De la Presa,
Rose Luckin,
Amanda Carr,
Thomas Dylan,
Abigail Durrant,
John Vines,
Shaun Lawson
Abstract:
New interactive physical-digital play technologies are sha** the way children plan. These technologies refer to digital play technologies that engage children in analogue forms of behaviour, either alone or with others. Current interactive physical-digital play technologies include robots, digital agents, mixed or augmented reality devices, and smart-eye based gaming. Little is known, however, a…
▽ More
New interactive physical-digital play technologies are sha** the way children plan. These technologies refer to digital play technologies that engage children in analogue forms of behaviour, either alone or with others. Current interactive physical-digital play technologies include robots, digital agents, mixed or augmented reality devices, and smart-eye based gaming. Little is known, however, about the ways in which these technologies could promote or damage child development. This systematic review was aimed at understanding if and how these physical-digital play technologies promoted developmentally relevant behaviour in typically develo** 0 to 12 year-olds. Psychology, Education, and Computer Science databases were searched producing 635 paper. A total of 31 papers met the inclusion criteria, of which 17 were of high enough quality to be included for synthesis. Results indicate that these new interactive play technologies could have a positive effect on children's developmentally relevant behaviour. The review indicated specific ways in which different behaviour were promoted. Providing information about own performance promoted self-monitoring. Slowing interactivity, play interdependency, and joint object accessibility promoted collaboration. Offering delimited choices promoted decision making. Problem solving and physical activity were promoted by requiring children to engage in them to keep playing. Four principles underpinned the ways in which physical digital play technologies afforded child behaviour. These included social expectations framing play situations, the directiveness of action regulations (inviting, guiding or forcing behaviours), the technical features of play technologies (digital play mechanics and physical characteristics), and the alignment between play goals, play technology and the play behaviours promoted.
△ Less
Submitted 10 February, 2022; v1 submitted 22 May, 2021;
originally announced May 2021.
