BACKGROUND People have insufficient knowledge and many misconceptions about the blood donation process, which hampers donors recruitment. Therefore, novel strategies and resources are needed to provide information and improve these… Click to show full abstract
BACKGROUND People have insufficient knowledge and many misconceptions about the blood donation process, which hampers donors recruitment. Therefore, novel strategies and resources are needed to provide information and improve these circumstances. OBJECTIVE We aimed at an interactive conversational agent to explain about blood donation. METHODS We used the Dialogflow framework to develop a conversational agent and deployed it publicly. Afterward, we conducted an assessment of user experience (UX) with 50 participants who interacted with the agent. We analyzed participants' opinions, the different UX scales, and their association with participants' demographic variables. RESULTS The conversational agent is available on the Google Assistant platform in Brazil. It is capable of responding to utterances related to 30 common questions and concerns about donating blood. The user can interact and explore freely and in any order by typing, speaking and selecting interface elements. The agent responds by speaking and displaying visual information, some multimedia content, and suggestions for continuing the dialogue. It enables a conversational sequence in which knowledge is imparted to the user in stages as the dialogue evolves. The overall UX assessed was very satisfactory, and people with specific demographic characteristics were more likely to have better UX. All participants had positive opinions and attitudes towards the conversational agent. CONCLUSIONS A conversational agent is a creative and captivating strategy of imparting knowledge and engage people regarding blood donation. The findings reaffirm the potential of using this technology for information outreach, especially for socially relevant purposes.
               
Click one of the above tabs to view related content.