Abstract
Chatbots systems, despite their popularity in today's HCI and CSCW research, fall short for one of the two reasons: 1) many of the systems use a rule-based dialog flow, thus they can only respond to a limited number of pre-defined inputs with pre-scripted responses; or 2) they are designed with a focus on single-user scenarios, thus it is unclear how these systems may affect other users or the community. In this paper, we develop a generalizable chatbot architecture (CASS) to provide social support for community members in an online health community. The CASS architecture is based on advanced neural network algorithms, thus it can handle new inputs from users and generate a variety of responses to them. CASS is also generalizable as it can be easily migrate to other online communities. With a follow-up field experiment, CASS is proven useful in supporting individual members who seek emotional support. Our work also contributes to fill the research gap on how a chatbot may influence the whole community's engagement.
Highlights
Chatbot systems1 have been increasingly adopted in many fields (e.g., healthcare [30], human resources (HR) [64], and customer service [122]), since the first chatbot system—ELIZA—emerged in 1964 to provide consulting sessions as a computer therapist [115]
We propose a neural-network based approaches (NN)-based chatbot architecture based on which a chatbot system can accurately handle unseen questions and generate various forms of responses with the same meaning
This paper proposes an example of deploying AI algorithms in an online community, but their algorithm is less explicit to the users, compared to the chatbot systems that we aim to develop and deploy in this paper
Summary
Chatbot systems have been increasingly adopted in many fields (e.g., healthcare [30], human resources (HR) [64], and customer service [122]), since the first chatbot system—ELIZA—emerged in 1964 to provide consulting sessions as a computer therapist [115]. Hu et al [44] built an experimental chatbot system that can understand the tones in a text input (e.g., sad or polite) and generate responses with an appropriate tone Following these system development efforts, many recent Human-Computer Interaction (HCI) and Computer-Supported Cooperative Work (CSCW) studies have examined various aspects of chatbots from the end users’ perspective, such as human-in-the-loop chatbot design [21], user perception of chatbots [18], playful usage of chatbots [64], and human trust in chatbots [49]. Inspired by previous literature [117], we build a human-in-the-loop module so a human operator can monitor and intervene the fully automated architecture, if needed
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
