Exploring Multi-Table Retrieval Through Iterative Search

TL;DR

This paper introduces an iterative search framework for multi-table retrieval in open-domain question answering, balancing relevance, coverage, and joinability, achieving high performance with significantly improved efficiency.

Contribution

It presents a novel iterative heuristic approach for multi-table retrieval that is scalable, interpretable, and effective, outperforming exact optimization methods in speed while maintaining competitive accuracy.

Findings

Achieves competitive retrieval performance with MIP-based methods.

Runs 4-400x faster depending on the benchmark.

Demonstrates effectiveness across 5 NL2SQL benchmarks.

Abstract

Open-domain question answering over datalakes requires retrieving and composing information from multiple tables, a challenging subtask that demands semantic relevance and structural coherence (e.g., joinability). While exact optimization methods like Mixed-Integer Programming (MIP) can ensure coherence, their computational complexity is often prohibitive. Conversely, simpler greedy heuristics that optimize for query coverage alone often fail to find these coherent, joinable sets. This paper frames multi-table retrieval as an iterative search process, arguing this approach offers advantages in scalability, interpretability, and flexibility. We propose a general framework and a concrete instantiation: a fast, effective Greedy Join-Aware Retrieval algorithm that holistically balances relevance, coverage, and joinability. Experiments across 5 NL2SQL benchmarks demonstrate that our iterative method achieves competitive retrieval performance compared to the MIP-based approach while being 4-400x faster depending on the benchmark and search space settings. This work highlights the potential of iterative heuristics for practical, scalable, and composition-aware retrieval.