Fairness-Aware Question Answering for Intelligent Assistants
Abstract: Conversational intelligent assistants, such as Amazon Alexa, Google Assistant, and Apple Siri, are a form of voice-only Question Answering (QA) system and have the potential to address complex information needs. However, at the moment they are mostly limited to answering with facts expressed in a few words. For example, when a user asks Google Assistant if coffee is good for their health, it responds by justifying why it is good for their health without shedding any light on the side effects coffee consumption might have \citegao2020toward. Such limited exposure to multiple perspectives can lead to change in perceptions, preferences, and attitude of users, as well as to the creation and reinforcement of undesired cognitive biases. Getting such QA systems to provide a fair exposure to complex answers -- including those with opposing perspectives -- is an open research problem. In this research, I aim to address the problem of fairly exposing multiple perspectives and relevant answers to users in a multi-turn conversation without negatively impacting user satisfaction.
Loading