Reproduce “OGDA: Ontology-Grounded Data Agents for Reliable Analysis over Heterogeneous Data Lakes” 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
Boyi Zhang, Hong Ting Tsang, Yisen Gao, Qiao Xiao, LI Yufei, Haoyu Huang, Zhongwei Xie, Jiaxin Bai and 1 more
OGDA builds reusable maps of record-structured data lakes and requires agents to ground selected files in a task graph before reading them. A proposer and verifier construct schemas and cross-file semantic links, while an experience store retains lessons. On KramaBench, OGDA scores 73.00% versus Reflexion’s 62.18%, though pipeline-functionality coverage gains are less consistent; it is designed for record-structured data.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.