Can You Describe Your Pain? Combining Psycho-emotional and Pragmatic Analysis on Cancer Patients’ Utterances

Abstract

This contribution is part of a wider research concerning Automatic Pain Assessment (APA), whose goal is to collect and analyse multimodal (audio and visual) data and to pave the way for more effective and personalized pain management strategies through innovative and multidimensional approaches. The APA project is dedicated to advancing innovative strategies grounded in artificial intelligence methodologies, with a central focus on probing pain features.

In this frame, the specific focus of this proposal is a pragmatic analysis of oncological patients’ utterances through the implementation of a modified version of the Kübler-Ross’s 5 grief stages model and a pragmatic classification scheme with 5 macro-types (assertion, direction, expression, rituals, dialogic moves) and, for each of these, a complex set of sub-type labels. A detailed analysis of the annotation results is presented, both separately and considering their relative covariation. The study provides a deeper understanding of an individual's pain experience. This, consequently, has practical importance as it facilitates the development of superior techniques to increase individual well-being and general health.