RedArchive+: A Human-AI Knowledge Retrieval System for Revolutionary Literature Co-Developed By Chinese Academy Of Social Sciences and Renmin University Of China

Abstract

RedArchive+ is a human-AI collaborative platform designed for the semantic retrieval of revolutionary literature. It has been jointly developed by the Institute of Modern History at the Chinese Academy of Social Sciences and the Library of Renmin University of China. The system integrates a restructured metadata model with three core modules: entity-level knowledge querying, context-aware semantic expansion, and interactive human-AI feedback facilitated through dynamic knowledge graphs.

When applied to key datasets in red literature, RedArchive+ significantly enhances retrieval accuracy, task relevance, and user satisfaction. Leveraging advanced technologies such as topic models, large language models (LLMs), and retrieval-augmented generation (RAG), RedArchive+ supports personalized access to historical texts that is semantically enriched. This advancement contributes to the ongoing research in digital humanities focused on red literature.