HELIOS: Harmonizing Early Fusion, Late Fusion, and LLM Reasoning for Multi-Granular Table-Text Retrieval

2026-02-25

AI Summary

Existing table-text retrieval methods suffer from the introduction of irrelevant context in early fusion and the risk of missing critical information in late fusion, while also struggling to support complex operations such as column aggregation and multi-hop reasoning. To address these limitations, this work proposes a unified retrieval framework that synergistically combines the strengths of both early and late fusion. The approach leverages edge-level bipartite subgraphs for fine-grained alignment, incorporates a query-driven dynamic subgraph expansion mechanism, and constructs a star-graph structure to guide large language models in hierarchical logical reasoning. Evaluated on the OTT-QA benchmark, the method significantly outperforms state-of-the-art approaches, achieving relative improvements of 42.6% in recall and 39.9% in nDCG, effectively balancing contextual relevance with advanced reasoning capabilities.

Technology Category

• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Search and Optimization: Learning to Search
• Data Mining & Knowledge Management: Intelligent Query Processing

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Table-text retrieval aims to retrieve relevant tables and text to support open-domain question answering. Existing studies use either early or late fusion, but face limitations. Early fusion pre-aligns a table row with its associated passages, forming "stars," which often include irrelevant contexts and miss query-dependent relationships. Late fusion retrieves individual nodes, dynamically aligning them, but it risks missing relevant contexts. Both approaches also struggle with advanced reasoning tasks, such as column-wise aggregation and multi-hop reasoning. To address these issues, we propose HELIOS, which combines the strengths of both approaches. First, the edge-based bipartite subgraph retrieval identifies finer-grained edges between table segments and passages, effectively avoiding the inclusion of irrelevant contexts. Then, the query-relevant node expansion identifies the most promising nodes, dynamically retrieving relevant edges to grow the bipartite subgraph, minimizing the risk of missing important contexts. Lastly, the star-based LLM refinement performs logical inference at the star graph level rather than the bipartite subgraph, supporting advanced reasoning tasks. Experimental results show that HELIOS outperforms state-of-the-art models with a significant improvement up to 42.6\% and 39.9\% in recall and nDCG, respectively, on the OTT-QA benchmark.

Problem

• table-text retrieval
• early fusion
• late fusion
• multi-hop reasoning
• column-wise aggregation

Innovation

• table-text retrieval
• early-late fusion
• bipartite subgraph
• LLM reasoning
• multi-granular retrieval