用大模型+知识图谱打造可对话的论文导航工具，让研究者能像聊天一样查文献。
Chatting with Papers: A Hybrid Approach Using LLMs and Knowledge Graphs
- 结合大模型与知识图谱，实现对论文集合的自然语言交互
- 支持研究者逐步明确问题，从概览到细节层层深入
- 适用于社科类论文集，可扩展至其他领域文献管理
本文介绍了一种名为GhostWriter的新工作流，将大语言模型与知识图谱（语义实体）结合，用于支持文献集合的导航。该工作流基于后端工具套件EverythingData构建，提供查询和“对话”式接口，使研究人员能以自然语言与论文集合互动。通过迭代使用，该系统可满足研究者在阅读文献时的信息需求：获取整体概览、深入理解特定概念及其上下文，并帮助其在可控范围内优化研究问题。我们以GESIS—莱布尼茨社会科学研究机构出版的《方法数据分析》期刊文章集合为例进行演示，同时指出该技术在其他领域的潜在应用价值。
原文摘要 · Abstract (English)
This demo paper reports on a new workflow \textit{GhostWriter} that combines the use of Large Language Models and Knowledge Graphs (semantic artifacts) to support navigation through collections. Situated in the research area of Retrieval Augmented Generation, this specific workflow represents the creation of local and adaptable chatbots. Based on the tool-suite \textit{EverythingData} at the backend, \textit{GhostWriter} provides an interface that enables querying and ``chatting'' with a collection. Applied iteratively, the workflow supports the information needs of researchers when interacting with a collection of papers, whether it be to gain an overview, to learn more about a specific concept and its context, and helps the researcher ultimately to refine their research question in a controlled way. We demonstrate the workflow for a collection of articles from the \textit{method data analysis} journal published by GESIS -- Leibniz-Institute for the Social Sciences. We also point to further application areas.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。