Exploration of Emotion Modelling Through Fuzzy Logic




Journal Title

Journal ISSN



Volume Title


De Montfort University


Thesis or dissertation

Peer reviewed


This work outlines a programme of research tasked with the exploration of representing psychologically grounded theories of emotion through fuzzy logic systems. It presents an introduction to the specific goals of the project, followed by an overview of the wider, multi-disciplinary field of emotion representation.

Two emotion theories are explored in detail. One, rooted in behaviourism, proposed by J. R. Millenson in 1967; the other, the Geneva Emotion Wheel proposed by K. R. Scherer in 2005. Each of these theories is independently abstracted mathematically, and represented in terms of both type-1 and type-2 fuzzy logic systems. Six potential implementations of these systems are presented. Of these, five are tested within this report. The results of these tests are analysed and discussed in the context of both computational behaviour and psychological analogue. There follows a critical review where the effectiveness of the different implementations and models is considered, informed by both testing results and the psychology upon which they are based.

A prototype of one implementation applied to govern the behaviour of an agent in a predator-prey scenario is included. Discussion of this prototype includes examples of how the implementation was practically applied to the environment, and an assessment of the behaviours of the agent in testing.

The work concludes with an overview of the thesis, including discussion of the results of the project and future avenues of research related to the completed work. The contributions of the thesis are explicitly outlined: the research of pre-existing, psychologically grounded models of emotional state suitable for computational representation; construction of mathematical representations of two models of emotion, using both type-1 and type-2 fuzzy logic; and, the presentation of five computational implementations of those representations, of which four are explicitly tested, compared and critically reviewed.



artificial intelligence, emotion modelling, Fuzzy Logic



Research Institute