Chatbot Design Challenges and the Effect on User Behavior

Abstract

Chatbots are efficient artificial intelligence tools employed frequently across various industries such as healthcare, tourism, entertainment, and education, to assist in performing repetitive activities requiring a conversation, thereby facilitating humans to focus on more innovative tasks. This chapter reviews the design challenges of two types of chatbots classified based on their interaction modes: dyadic chatbots interacting with one individual at a time; and polyadic chatbots interacting with multiple individuals simultaneously. Further, the chapter identified the associated challenges in the chatbots, namely: engagement, trust, and human likeliness based on personality types - at a higher level in the chatbot design, together with the impact of each challenge on the user behavior. The authors review the literature in the relevant areas to pinpoint the research gaps requiring more focus within each area of challenge. The research community can address these identified research areas, which will eventually promote a more efficient human-machine collaboration.