Search for: All records

Clinical group bereavement therapy often promotes narrative sharing as a therapeutic intervention to facilitate grief processing. Increasingly, people turn to social media to express stories of loss and seek support surrounding bereavement experiences, specifically, the loss of loved ones from suicide. This paper reports the results of a computational linguistic analysis of narrative expression within an online suicide bereavement support community. We identify distinctive characteristics of narrative posts (compared to non-narrative posts) in linguistic style. We then develop and validate a machine-learning model for tagging narrative posts at scale and demonstrate the utility of applying this machine-learning model to a more general grief support community. Through comparison, we validate our model's narrative tagging accuracy and compare the proportion of narrative posts between the two communities we have analyzed. Narrative posts make up about half of all total posts in these two grief communities, demonstrating the importance of narrative posts to grief support online. Finally, we consider how the narrative tagging tool presented in this study can be applied to platform design to more effectively support people expressing the narrative sharing of grief in online grief support spaces. 

Schema induction builds a graph representation explaining how events unfold in a scenario. Existing approaches have been based on information retrieval (IR) and information extraction (IE), often with limited human curation. We demonstrate a human-in-the-loop schema induction system powered by GPT-3. We first describe the different modules of our system, including prompting to generate schematic elements, manual edit of those elements, and conversion of those into a schema graph. By qualitatively comparing our system to previous ones, we show that our system not only transfers to new domains more easily than previous approaches, but also reduces efforts of human curation thanks to our interactive interface.