InfoGatherer: Principled Information Seeking via Evidence Retrieval and Strategic Questioning

TL;DR

InfoGatherer is a framework that improves information gathering for high-stakes question-answering by systematically combining evidence from documents and user interactions, leading to more reliable and interpretable decisions.

Contribution

It introduces a principled uncertainty modeling approach using Dempster-Shafer theory to fuse evidence from multiple sources in LLM-based QA systems.

Findings

Outperforms baselines on legal and medical tasks

Requires fewer interaction turns for accurate answers

Provides more trustworthy and interpretable outputs

Abstract

LLMs are increasingly deployed in high-stakes domains such as medical triage and legal assistance, often as document-grounded QA systems in which a user provides a description, relevant sources are retrieved, and an LLM generates a prediction. In practice, initial user queries are often underspecified, and a single retrieval pass is insufficient for reliable decision-making, leading to incorrect and overly confident answers. While follow-up questioning can elicit missing information, existing methods typically depend on implicit, unstructured confidence signals from the LLM, making it difficult to determine what remains unknown, what information matters most, and when to stop asking questions. We propose InfoGatherer, a framework that gathers missing information from two complementary sources: retrieved domain documents and targeted follow-up questions to the user. InfoGatherer models uncertainty using Dempster-Shafer belief assignments over a structured evidential network, enabling principled fusion of incomplete and potentially contradictory evidence from both sources without prematurely collapsing to a definitive answer. Across legal and medical tasks, InfoGatherer outperforms strong baselines while requiring fewer turns. By grounding uncertainty in formal evidential theory rather than heuristic LLM signals, InfoGatherer moves towards trustworthy, interpretable decision support in domains where reliability is critical.