Computing Emotion Dynamics from Text and Exploring their use as Biosocial Markers of Overall Well-being

• Author / Creator

  Teodorescu, Daniela

• Language is inherently social -- it is influenced by our lived experiences and environments, and impacts the way in which we communicate with each other. Therefore, it is not a surprise that our health impacts our language. The patterns with which emotion changes over time -- emotion dynamics -- is a framework in psychology that is crucial for better understanding overall well-being and mental health. While emotion dynamics (ED) are commonly measured through self-report surveys, recent work has attempted to mitigate limitations of surveys (e.g., limited reach, bias, etc.) through inferring ED from utterances. The utterance emotion dynamics (UED) framework consists of four commonly used metrics, and we expand it by computing two additional metrics inspired by ED work in psychology, for a total of six metrics.

  We investigate the relationship between UED and well-being in two domains -- poems written by children across grades and data related to mental health. We compute UED metrics on poems written by children in grades 1 to 12 across seven emotions and find trends of emotion change across grades. For valence, these patterns of emotional change across grades are supported by previous work in psychology. Here, we further explored two other dimensions of emotions (arousal and dominance) and four categorical emotions (anger, fear, joy, and sadness), further contributing novel results for these emotions. We extend this work to characterize the UED of groups with a mental health diagnosis on Twitter. Overall, our work builds upon the UED framework by computing new metrics and applying it to various domains. We demonstrate that patterns of emotional change in text can act as an indicator of overall well-being across domains.

• Subjects / Keywords

  • Utterance Emotion Dynamics
  • Sentiment Analysis
  • Emotion Lexicons
  • Emotion Dynamics
  • Well-Being
  • Mental Health
  • Children's Writing
  • Poems
  • Social Media
  • Twitter
  • Emotion Arcs
  • Valence
  • Sentiment
  • Children’s Writing
  • Children’s Poems

• Graduation date

  Fall 2023

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/r3-xc46-yv25

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.