Dayan Pan, Bincheng Zuo, Xiaojian Wang, Daogang Zhang, Jingyuan Wang
AgentOnto is an ontology-guided LLM agent for enterprise data analytics that organizes heterogeneous knowledge into an ontology to guide analysis. It incorporates confirmed feedback and improves through iterative ontology preparation. This approach supports reliable and adaptable data analysis.
AgentOnto's ability to explicitly construct and update enterprise knowledge enables reliable data analysis, which is crucial for informed decision-making in enterprises. Its practical value is demonstrated through a real-world deployment, showcasing its potential to improve enterprise analytics. This can lead to better business outcomes and more efficient data-driven decision-making.
Thank you to arXiv for use of its open access interoperability. arxiv.org. aibrief.ai is not affiliated with or endorsed by arXiv.