The success of user interfaces based on conversational agents is very much dependent on the agents' conversational skills, and their ability to consistently portray personality and react in an emotionally plausible way in a social context. In the NECA project, the focus is on the design of credible agent-agent interaction patterns to be observed by human users. To achieve a high level of credibility, the agents must be able to express themselves using a combination of verbal and non-verbal output driven by personality and emotion models.
The NECA project will develop a new generation of mixed multi-user / multi-agent virtual spaces populated by affective conversational agents. The agents will be able to express themselves through synchronised emotional speech and non-verbal behaviour, generated from an abstract representation which can be the output of an affective reasoner. This is the first time that such expressive capabilities are featured in Internet applications. The agents' usefulness will be evaluated in two concrete application scenarios. From a technical point of view, the emerging NECA platform will provide a confederation of dedicated components including an affective reasoner, co-ordinated generation, and emotional speech synthesis, thus providing a basis for the development of new Internet applications with emotional agents.
The main contribution of DFKI comprises, on the one hand, the automatic assessment of a situation with the aid of personality and emotion models, and on the other hand, the enhancement of existing speech synthesis with means for expressing emotions.
The demonstrators developed in the project will be made publicly accessible on the World Wide Web.
Partners
- ÖFAI, Wien (Koordinator)
- DFKI GmbH, Saarbrücken
- Freeserve Plcm Hertfordshire, UK
- Univerdsität des Saarlandes, Saarbrücken
- Sysis Interactive Simualtions Ag, Wien
- University of Brighton, Brighton, UK
