Enhancing user's self-disclosure through chatbot's co-activity and conversation atmosphere visualization

Abstract

Prior research showed that chatbot had a great potential to elicit users' self-disclosure because it is not judging its users. However, chatbot’s feature other than conversation’s effect to elicit user’s self-disclosure is still understudied. In this study, we developed a chatbot and implemented two features other than conversation, which are co-activity (COA) and conversation atmosphere visualization (CAV) to elicit user’s self-disclosure. We conducted a field study involving 87 participants who were randomly assigned to four different groups (control, COA only, CAV only, CAVCOA) for 10 days. Our results show that both COA and CAV features positively affect user’s objective self-disclosure. In addition, interaction effects between COA and CAV have been found to affect user’s perceived intention to use and enjoyment. We also provide discussions on how COA and CAV should be designed to improve user's relationship development with the chatbot and to promote user’s self-disclosure.