Paperlib is an academic literature management and retrieval server designed to streamline research workflows. It facilitates the intelligent import of PDF documents, automatically extracting text, chunking content by page, and generating vector embeddings for sophisticated search. Users can leverage robust hybrid search capabilities that combine full-text and semantic queries for comprehensive information retrieval. Furthermore, the tool excels in knowledge synthesis, employing LLMs to build dynamic knowledge graphs with entity, relation, and claim extraction, along with community detection, culminating in the ability to auto-generate structured literature reviews based on curated evidence packs.
主要功能
01LLM-driven knowledge graph construction with entity/relation/claim extraction
02Automatic PDF text extraction and vector embedding generation
03Hybrid search combining full-text and pgvector semantic search
04Leiden community detection for topic analysis within knowledge graphs
05Automated structured literature review generation from evidence packs
060 GitHub stars
