A Flexible Mapping Scheme for Discrete and Dimensional Emotion Representations


While research on emotions has become one of the most productive areas at the intersection of cognitive science, artificial intelligence and natural language processing, the diversity and incommensurability of emotion models seriously hampers progress in the field. We here propose kNN regression as a simple, yet effective method for computationally mapping between two major strands of emotion representations, namely dimensional and discrete emotion models. In a series of machine learning experiments on data sets of textual stimuli we gather evidence that this approach reaches a human level of reliability using a relatively small number of data points only.

Back to Table of Contents