Advances in Human-Computer Interaction / 2014 / Article / Fig 1

Research Article

A Hierarchical Probabilistic Framework for Recognizing Learners’ Interaction Experience Trends and Emotions

Figure 1

The proposed framework for assessing learners’ emotions and interaction experience trends using a DBN. The rectangular spaces are generic; other variables can be inserted or replaced given the available modalities. For instance, the sensory nodes can be adapted to the current environment and devices (e.g., a video camera, a posture sensitive chair, an eye tracker, etc.). Dashed arrows denote temporal dependencies; for example, the interaction trend at time is affected by the experience trend at time . Note that we do not draw all the links between the emotion nodes and each diagnostic variable due to space limit.