Natural Language Dialogue group

We define a question-answering character as a virtual persona who answers questions in character, as if being interviewed by a human user. Of course there are many types of interview (job interview, reporter interview, police interview, ...), as well as questioning in casual conversation. Moreover, different people and personality types react differently to being interviewed. A designer of a question-answering focuses on the character's response appropriateness. This property distinguishes question-answering characters from question-answering systems, where both questions and answers are based on facts and the goal is to provide the most relevant answer. For example, depending on the scenario, an evasive, misleading, or an “honestly” wrong answer from a question-answering character might be appropriate but would not be relevant.