Integration von emotionaler Intelligenz in Interface-Agenten am Beispiel einer Trainingssoftware für lernbehinderte Kinder

Adaptive User Interfaces are an important success factor for the development of information and communication systems and, in particular, for those subsystems where interaction with end-users is critical, like therapeutical and training systems. Current adaptive interfaces are based upon user and device profiles. The presented work adds a new powerful tool for the development of such adaptive systems by analysing the arousal states of the user with non-invasive biofeedback sensors. This provides a reliable mean to recognise the arousal states of the user that complements other research attempts in this area that interpret arousal states by other external signals of the human body, like gestures, speech intonation, blood pressure, skin-conductivity etc.. Based upon this, architecture to build generic emotional computer systems was constructed, which could be used as a template to create such systems in different application areas. The developed system proposes a combination of sensors and a set of effective algorithms to infer the arousal state of the user. This information is used to modify the behaviour and the interfaces of the application. In particular, a therapeutical system was developed where Interface Agents – which themselves can communicate through mimic, speech and affective expressions with the user – modified their behaviour according to the arousal status of the user. This approach can be extended to develop narrative training and therapeutical systems, which could adapt their functionality and user interface to the motivation level of the user, and focus her attention on the fulfilment of the current task. The aforementioned recognition process elicits the user’s arousal states at a level sophisticated enough for the target systems. It is beyond the scope of this work to imitate affective or recognition abilities of human beings. The process describes several input devices to measure the arousal state of a user and shows a concept for its interpretation. The architecture consists of the following components: 1. Input devices, e.g., optical sensors, acoustical sensors or biofeedback sensors. 2. Assessment methodologies and stochastic methods. 3. Simulation of emotions in Interface Agents. 4. Integration in applications, and especially as a multimodal Human-Computer Interface. As a proof-of-concept, the author developed and thoroughly evaluated a therapeutical system – TAPA, Training by Animated Pedagogical Agents – targeted to learning disabled children. These children have frequently deficits in their meta-memory strategies, i.e., the users’ knowledge about their own memory abilities, like categorising objects. TAPA was based upon the concepts described above and used the skin conductivity sensor as the input device to measure the user’s arousal states. The system trains end-users to improve their meta-memory abilities, and the evaluations performed with groups of children with a variety of learning disabilities have shown very outstanding results, as children who used the system greatly improved their skills in this area in comparison to different control groups. The building of TAPA, and in general of any emotional computer system, requires a multidisciplinary approach where different disciplines and technologies collaborate together, especially computer science and psychology. Among the utilised technologies in this work, we can highlight: interface design and story-telling, user modelling, e-learning and biofeedback sensors.

Grantor dc:publisher

Publikationsserver der RWTH Aachen University

Year dc:date

2005

Author dc:creator

• Mohamad, Yehya

Contributors dc:contributor

• Berlage, Thomas

• info:eu-repo/classification/ddc/004
• Multimodales System
• Mehragentensystem
• Gefühl
• Benutzerprofil
• Kind
• Lernbehinderung
• Informatik
• Emotion
• Sensoren
• Interface
• Agenten
• HCI

Statement dc:rights

• info:eu-repo/semantics/openAccess

Language dc:language

ger

Identifier

https://publications.rwth-aachen.de/search?p=id:%22RWTH-CONV-123707%22

OAI identifier oai:identifier

oai:publications.rwth-aachen.de:62114