The purpose of this research is to examine the implementation of an emotionally intelligent computing system, and possible applications of such a system. In theory, if computing systems are better able to "understand" the emotions of people, then people will be more motivated to communicate with the system. This type of research has relevance in many different areas, both in academia and industry. Programming emotional intelligence will improve communication between humans and devices ranging from customer service chatbots, to cell phones, to cars.
The study is attempting to explore current research in the field, as well as expand upon Professor Clark Elliott's "Affective Reasoner" implementation. Of particular interest is the categorization of the actions of other agents, and the manifestation of emotions based on this categorization. This research attempts to generate a method for answering the following questions:
The research focuses on methodology for gathering, indexing, and mapping input for emotion eliciting situations with the purpose of creating a "compassionate" computing system.
A Simulation Event, in this context, is a representation of many factors arising from a life event that may affect the emotion(s) of the user. This model considers time to unfold as a series of discrete and separate events. While some simulation events within the original Affective Reasoner Model do not have any effect on emotion generation, our research is concerned with emotion eliciting situation frames. These events are used in the Affective Reasoning system to elicit an emotional response from the intelligent agents.
Emotion eliciting situation frames can be either physical events, or more abstract occurrences like dreams or thoughts. These situations can change the overall state (affect) of the system, or could even trigger a response (action). Additional description of Emotion Eliciting Condition Relations as described by Professor Elliott can be found here.