Examining Usability on Atreya Bot: A Chatbot Designed for Chemical Scientists

Conversational agents are a contemporary topic in human-computer interaction, implemented in multi-disciplinary industrial applications to support consumers. Information seeking task can be competently incorporated into dialogue-based agents, so that searchers can easily fulfil their information needs. In this study, we proposed "Atreya", an interactive chatbot for chemistry scholars to examine and access the ChEMBL database. Atreya chatbot accesses the ChEMBL database and extracts drug-related information such as relevant drugs for a particular disease and associated targets. Telegram, a cloudbased messaging application is used to host Atreya. In this study, respondents were given a complex information seeking task to complete in order to conduct usability tests on Atreya Bot. The usability questionnaire based on different metrics was filled out by the searchers during information seeking activity on the Atreya Bot. The results were then calculated, and analysis was performed by 3 standard usability metrics - SUS (System Usability Scale), CUQ (Chatbot Usability Questionnaire) and UEQ (User Experience Questionnaire). These questionnaires help us measure attributes such as usability, efficiency, etc., provided by the bot. We found positive feedback results for all three metrics, which mainly emphasize that the Atreya bot is simple to use as the numerous functions in this system are well integrated and can support students in information seeking.