LUFY-Dataset

该数据集名为LUFY-Dataset，包含了从用户实验中收集的对话，这些实验涉及三种类型的RAG聊天机器人：原始RAG、MemoryBank和LUFY。数据集包括参与者创建的612个问答对，这些问答对根据记忆的重要性进行了标注。此外，该数据集为长期对话交流提供了基准，并通过主观和自动化的方法评估用户体验。规模上，数据涵盖了17名参与者在4天内进行的10个会话的数据。该数据集的任务是评估RAG聊天机器人在长期对话和记忆保持方面的表现。

This dataset, named LUFY-Dataset, contains conversations collected from user experiments involving three types of RAG chatbots: vanilla RAG, MemoryBank, and LUFY. It includes 612 question-answer pairs created by participants, which are annotated based on the importance of the involved memories. Additionally, this dataset provides a benchmark for long-term conversational interactions, and evaluates user experience through both subjective and automated methods. In terms of scale, the dataset covers data from 10 sessions conducted by 17 participants over a 4-day period. The task of this dataset is to evaluate the performance of RAG chatbots in long-term conversations and memory retention.