Conversational Interfaces for Explainable AI: A Human-Centered Approach

One major goal of Explainable Artificial Intelligence (XAI), in order to enhance trust in technology, is to enable the user to enquire information and explanation about its functionality directly from an intelligent agent. We propose conversational interfaces (CI) to be the perfect setting, since they are intuitive for humans and computationally processible. While there are many approaches addressing technical issues of this human-agent communication problem, the user perspective appears to be widely neglected. With the purpose of better requirement understanding and identification of implicit expectations from a human-centered view, a Wizard of Oz experiment was conducted, where participants tried to elicit basic information from a pretended artificial agent (What are your capabilities?). The hypothesis that users pursue fundamentally different strategies could be verified with the help of Conversation Analysis. Results illustrate the vast variety in human communication and disclose both requirements of users and obstacles in the implementation of protocols for interacting agents. Finally, we infer essential indications for the implementation of such a CI.