Multimodal recognition of affect over the course of a tutorial learning experiment

辅导学习实验过程中情感的多模态识别

• 批准号: 394413900
• 负责人: Dr. André Brechmann
• 金额: --
• 依托单位: Institut für Neuroinformatik (NI)
• 依托单位国家: 德国
• 项目类别: Research Grants
• 财政年份: 2018
• 资助国家: 德国
• 起止时间: 2017-12-31 至 2021-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/394413900?language=en
• 关键词: Multimodal; recognition; affect; over; course

Tutorial systems in a learning environment should support the user in reaching his specific learning goals and react to his level of knowledge and individual abilities in a pedagogical reasonable manner. For this a comprehensive model of the user is essential taking the prior knowledge and interaction history of the user into account as well as his current affective state. In our research project we will focus on an interactive learning task, in which the user is supported by a tutorial system component. Neurophysiological data (fMRT, EEG) and psychophysiological data (ECG, skin conductance, respiration) as well as details of the behavior of the user (dynamics of key stroke, facial expressions) are recorded synchronously, in order to determine changing affective and cognitive user states by multimodal data analysis. The interdisciplinary analysis and interpretation of the interconnection between the partially weak signals of the single modalities, and the development and optimization of the classifiers for affect recognition are the primary goals of our research project.

在学习环境中，学习系统应该支持用户达到他的特定学习目标，并以合理的教学方式对他的知识水平和个人能力做出反应。为此，用户的综合模型是必不可少的，考虑到用户的先验知识和交互历史以及他当前的情感状态。在我们的研究项目中，我们将专注于一个互动的学习任务，其中用户支持的教程系统组件。同步记录神经生理学数据（fMRT、EEG）和心理生理学数据（ECG、皮肤电导、呼吸）以及用户行为的细节（击键动态、面部表情），以便通过多模态数据分析确定不断变化的情感和认知用户状态。跨学科的分析和解释之间的互连的部分弱信号的单一模态，以及发展和优化的分类器的影响识别是我们的研究项目的主要目标。