Applications of Digital Hermeneutics to Suicide Prevention and Digital Mental Health

Digital hermeneutics can be understood as involving interpreting verbal or non-verbal communication with machines, and also equipping machines in their endeavours to interpret. Approaches include (i) what are the scientific methods for digital hermeneutics?, (ii) what are the scientific questions for digital hermeneutics? This presentation examines the scientific methods for digital hermeneutics in the context of two applied issues, one within the realm of suicide prevention, one within the realm of digital mental health. One issue within the realm of suicide prevention is how we determine media adherence to Samaritans guidelines. Research evidence shows that certain types of media depictions can lead to imitational suicidal behaviour among vulnerable people. The Samaritans have developed media reporting guidelines. This brings the challenge of monitoring how all media reports adhere to these guidelines. The challenge of designing a coding framework which allows digital technologies to code large sources of data in a way which is accurate and considers both standardizations and nuances is part of digital hermeneutics. Certain organisations do try to manually code how media text adheres to Samaritans guidelines, but it requires extensive human hours. Also, different researchers use different coding frameworks. How can machine be equipped in their endeavours to interpret text, and what scientific methods can digital hermeneutics use to develop standardized coding frameworks? This presentation discusses why standardized coding frameworks are important in how we interpret language and record statistics e.g. to allow trends to be monitored across time and settings through the application of machine learning techniques. Corresponding methods are discussed, with recent strives towards standardized psychological autopsies as an example. Consideration is then given to how these methods would be applied to develop a coding framework to assess adherence of media reports to Samaritans guidelines within a digital platform. Such standardized coding frameworks are essential to equipping machines to interpret data. The second question also relates to equipping machines in their endeavours to interpret, but it looks within the context of digital mental health and asks how we can equip digital interventions such as chatbots in their endeavours to provide satisfactory solutions to the challenges presented by users. Chatbots show great potential efficacy, but this is compromised by low user engagement. Many factors may underlie this, but user satisfaction is one component. From psychological literature, we know that there are many different approaches to coping, and certain approaches are better or less suited to different challenges. This presentation discusses the methods that digital hermeneutics might employ to ensuring that the solutions posited by chatbots are actually those which are most applicable to the situation. This incorporates ensuring that a database of evidence-based solutions underlies the content development of chatbots. The presentation outlines what such an evidence based database or framework of approaches to coping might look like and considers how such a database might be populated, with clinician input, ChatGPT and Natural Language Processing (NLP) all considered.