Reproduce “Conversational Exploration of Literature Landscape with LitChat” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Mingyu Huang, Shasha Zhou, Yuxuan Chen, Ke Li
LitChat proposes a system for conversational exploration of scientific literature that combines Large Language Models (LLMs) with data-mining techniques and a bibliographic knowledge graph. This approach enables comprehensive, objective, open, and transparent analysis of large-scale publication datasets, addressing limitations of pure LLM-based and manual review methods.
Key insights
- Combining conversational AI (LLMs) with structured data-mining techniques and a bibliographic knowledge graph (BKG) allows for comprehensive, objective, open, and transparent (COOT) literature exploration.
- A human-in-the-loop interactive query refinement process enhances the relevance and comprehensiveness of the initial literature search.
- Leveraging a BKG and a data-mining toolbox ensures that generated insights are data-backed and verifiable, mitigating LLM hallucination issues.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.