Source record

Published by arXiv cs.HC (T1 source). The original is at https://arxiv.org/abs/2610.11025.

Pipeline notes

The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.

Paper presenting DAG-EDA, a system where analysts and an LLM co-navigate exploratory data analysis via an intent graph that decomposes questions into concrete tasks while preserving alternative paths, and a multi-layered knowledge graph that externalizes the LLM's domain knowledge. Includes a usage scenario and a user study design.

A concrete research system that structures how LLMs participate in exploratory analysis, keeping alternative paths open and grounding model suggestions in inspectable structures rather than free-form text.

Cited by

No citations on record.