Conversational Moderation Evaluation Dataset

该数据集包含了已完成对话和调查，旨在评估对话式人工智能模型在在线讨论中作为调解者的有效性。为了确保多样性，参与者被限制在50个对话会话内。互动设计使得参与者无法意识到他们是在与机器人还是人类进行对话。每个调解机器人评估的对话数量为60个，参与者群体具有多样性。该任务的目标是评估对话式对话模型作为调解者的表现。

This dataset comprises completed dialogues and surveys, aiming to evaluate the effectiveness of conversational AI models as mediators in online discussions. To ensure diversity, participants were limited to 50 conversation sessions. The interaction was designed so that participants could not discern whether they were conversing with a robot or a human. Each mediating robot was assigned 60 conversations to evaluate, with a diverse participant cohort. The goal of this task is to assess the performance of conversational AI models serving as mediators.