Academic journal article Educational Technology & Society

Identification of Action Units Related to Affective States in a Tutoring System for Mathematics

Academic journal article Educational Technology & Society

Identification of Action Units Related to Affective States in a Tutoring System for Mathematics

Article excerpt

Introduction

A facial expression is a fundamental component of non-verbal human communication. It not only conveys responses to external stimuli but also motivates action and adds meaning and richness to human experiences. Affective computing (Picard, 2000) is a branch of computer sciences aimed at endowing technology with mechanisms for the recognition of, understanding of and reaction to human emotions. Emotions encompass a number of psychological and physiological reactions used as cues by others in non-verbal communication. These cues are a fundamental part of human communication and provide the rudiments to develop technology to detect emotions automatically.

The recognition of emotions is an open problem in the computer sciences (El Kaliouby & Robinson, 2005). Emotions influence all human activities, and the learning context whether in the classroom or with educational technology is no exception. Previous works have employed technology to recognize emotions by processing facial expressions, speech and other physiological signals (Shen et al., 2009). Of particular relevance to this investigation is the work of Craig et al. (2004) as it provides indications of the importance of emotions in the learning process mediated by educational technology. Craig et al. (2004) distinguished between emotions expressed in non-classroom settings such as joy or sadness and the emotions displayed during the learning process mediated by educational technology. As a consequence, Craig et al. (2004) proposed that only a sub-set of emotions is present in the educational context.

Following this distinction and with the aim of distinguishing between the bigger set of emotions and those specific to educational technology, the term "affective states" is employed in this paper to refer to boredom, frustration, confusion and cognitive engagement or flow (Craig et al., 2004). Although the expression of affect by students is a normal part of their interaction with educational technology, they could be more inclined to stronger affective reactions in situations involving problem-solving activities. In these contexts, educational technology is designed to encourage the resolutions of mathematical problems accompanied by cycles of mistakes and recovering from these mistakes. During these cycles, students could experience one or all the affective states as the activities proposed by the tutoring systems often offer challenging activities (D'Mello et al., 2010). To fully endow educational technology with affective computing capabilities it is necessary to develop tools to recognize, understand and react to students' affect. The recognition of affective states has remained elusive and focused on physiological and psychological aspects related to affect. Previous research on affect has mainly been directed at the recognition problem because of the difficulties and ambiguities associated with the recognition of human emotions (Ocumpaugh et al., 2012). Some progress has been made, but a reliable, autonomous and context-independent solution to recognize affect during the use of educational technology does not exist. One of the problems with the development of such a tool is the context in which recognition takes place.

This paper assumes that cultural differences among students (Terzis et al., 2013) add difficulties to the development of a generalized affect detector. This paper deals with the problem of recognition of affective states, aiming at providing evidence that the employment of action units (conceived as universal facial reactions) could aid the recognition and understanding of students' affective states based on facial movements. The findings presented in this paper also support the idea that students undergo affective trajectories and suggest these have impact on the students' learning. This paper also offers pointers in relation to affective display in a Mexican secondary school but does not elaborate on how these affective displays compare to students with a different cultural background. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.