Evaluating Domain-Independent Small Talk Conversations to Improve Clinical Communication Interaction for Human and Machine.

Conversation as trivial as “small talk” has a significant impact on human relationships, including in clinical care environments. Evidence suggests that small talk can improve health and clinical outcomes. This study aims to analyze the characteristics of domain-independent small talk collected from a sample of conversations. We reviewed the impact of personality factors of the conversation and present our findings on specific factors that are correlated with it. In addition, we discuss the possibility of integrating small talk into our existing ontology-driven dialogue systems, which can assist machines in engaging with and building trust with patients in the clinical environments. Future direction will involve implementing the ontology model and integrating it with a health dialogue engine to test the effectiveness of automated small talk.