Fine-Grained Table Retrieval Through the Lens of Complex Queries

2026-03-07

AI Summary

This work addresses the challenge of efficiently retrieving relevant tables from relational databases for complex natural language questions. To tackle the difficulties posed by highly compositional queries and densely interconnected schemas in open-domain settings, the authors propose a fine-grained table retrieval method (DCTR) that leverages typed query decomposition and a global join-aware mechanism. The approach introduces a complexity-driven evaluation framework for retrieval and demonstrates significant performance gains over existing methods on an industrial-scale benchmark. Notably, DCTR exhibits superior robustness and accuracy in scenarios involving high query compositionality and dense join connectivity, effectively bridging the gap between natural language understanding and structured database querying.

Technology Category

Data Mining & Knowledge Management: Intelligent Query ProcessingKnowledge Representation and Reasoning: Computational Complexity of ReasoningNatural Language Processing: Question Answering

Application Category

Search and Retrieval-Augmented AI: Web evaluation methodologies and metricsGraph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphsWeb Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Enabling question answering over tables and databases in natural language has become a key capability in the democratization of insights from tabular data sources. These systems first require retrieval of data that is relevant to a given natural language query, for which several methods have been introduced. In this work we present and study a table retrieval mechanism devising fine-grained typed query decomposition and global connectivity-awareness (DCTR), to handle the challenges induced by open-domain question answering over relational databases in complex usage contexts. We evaluate the effectiveness of the two mechanisms through the lens of retrieval complexity which we measure along the axes of query- and data complexity. Our analyses over industry-aligned benchmarks illustrate the robustness of DCTR for highly composite queries and densely connected databases.

Problem

• table retrieval
• complex queries
• relational databases
• natural language question answering
• retrieval complexity

Innovation

• fine-grained table retrieval
• typed query decomposition
• global connectivity-awareness
• complex query
• relational database