Social Agents for Learning in Virtual Environments
Summary
It is imperative that health-care professionals are able to communicate well
with their patients. Not only verbally, but also in body language. They
should show empathy when needed and really know what to say and when
to say it. Previous efforts in this work by Jeuring et al. Communicate! -
A serious Game for Communication Skills [20] inspire us to expand upon
their work. Communicate! gives the ability to expand ones communication
skills, but does so in a restrictive manner. The player makes choices from a
limited set of responses. In this paper we present our application: SALVE
(Social Agents for Learning in Virtual Environments). SALVE is a serious
game that helps the player to develop their communication skills, but with
less restriction. The player speaks to an agent in a controlled environment,
but is free to speak as they wish. The agent responds based on pattern matching
via chat bot technology, assisted by a rule-based interpretation
module. Because of the endless possibilities that this combination provides,
a social practice is put in place to give a natural boundary to the scope of
the system. SALVE offers the ability to have a dialogue with an agent, that
shows emotion and accepts empathy. At the end, the player receives a score
based on how they did. Development in SALVE is easy because both the
pattern-matching and the rule-base languages are close to natural language.
In future iterations we hope to improve upon the ease of use of SALVE, as
well as the quality of the agents.