proFILL：面向大学学术咨询的画像感知检索增强聊天机器人
原标题：hoBIT: A Profile-Aware Retrieval-Augmented Chatbot for University Academic Advising
AI 摘要
本文介绍了 proFILL 方法，用于将大学学术咨询聊天机器人 hoBIT 从基于规则的对话系统升级为支持用户画像的检索增强生成（RAG）系统。proFILL 通过离线构建画像条件化索引，并在在线查询时按需获取用户画像属性，从而为不同院系、入学年份和专业的学生提供准确的答案。实验和用户偏好研究表明，proFILL 优于多种 RAG 基线，并能在开源模型中保持有效性，支持本地化部署。
正文节选
hoBIT: A Profile-Aware Retrieval-Augmented Chatbot for University Academic Advising Abstract In university academic advising, identical questions can require different answers depending on a student’s department, admission cohort, and degree program, causing profile-blind retrievers to surface plausible but inapplicable evidence. We present proFILL, a method for transforming hoBIT, our college’s current rule-based advising chatbot, into a profile-aware retrieval-augmented generation (RAG) system