Towards detecting the level of trust in the skills of a virtual assistant from the user’s speech

Abstract

Research has shown that trust is an essential aspect of human-computer interaction directly determining the degree to which the person is willing to use a system. Automatic detection of the level of trust that a user has in a certain system could be used to enable adaptive responses to correct potential distrust. We aim to explore the feasibility of automatically detecting whether the user trusts the ability of the virtual assistant (VA). We developed a novel protocol for collecting speech data from subjects interacting with VAs with different skill levels. The subject is asked to respond to a series of factual questions while interacting orally with the VA. It is presented as having been previously rated by other users as either competent or incompetent and it answers the subjects' questions consistently to its ability. The goal of the protocol was to induce subjects to either trust or distrust the VA's skills, assuming that this would affect their speech patterns. We collected a speech corpus in Argentine Spanish which is publicly available for research use. Using the collected data, we developed a system to detect the ability of the VA with which a subject interacted during a session, based on the subject's speech patterns, with an accuracy up to 76%, compared to a random baseline of 50%. Our analysis suggests that subjects change the way they speak to the VA depending on whether they perceive it as more or less competent; that is, depending on whether they trust its ability.