Creating a Chatbot for and with Migrants
Zhifa Chen, Yichen Lu, Mika Nieminen, Andrés Lucero
Abstract
Information portals are usually created to support the integration of migrants into a host country. However, the information-seeking process can be exhausting, cumbersome and even confusing for migrants as they must cope with time-consuming information overload while searching desired information from lists of documents. Chatbots are easy-to-use, natural, and intuitive, and thus could support information-seeking. There is a lack of research that engages and empowers migrants and other stakeholders as co-design participants in chatbot development. We explored how migrants can be empowered in designing a chatbot that supports their social integration. Using a co-design approach, we conducted a series of activities with migrants and other stakeholders (i.e., online questionnaires, empathy probes, surveys, and co-design workshops) to first understand their expectations regarding chatbots, and then co-design a personality-driven chatbot. We found that chatbot personality can drive co-designing a chatbot as design goals, design directions, and design criteria.