Correspondences as Decisions: JevNexus for Decision-Centric Schema Matching

Runze Li, Hanchen Wang, Ying Zhang, Wenjie Zhang
10/7/2026
cs.DBcs.AIcs.LG

Official code: RazeenLI/JevNexus (1 star, MIT, last commit 2026-10-07).

Abstract

Schema matching increasingly uses generative language models to rerank retrieved column candidates, although the underlying task is a bounded correspondence decision. We present JevNexus, which combines typed pairwise decisions with schema/instance evidence and invokes listwise refinement only when the evidence disagrees and the fused margin is small. The evaluation covers 561 cases from six benchmark families. JevNexus obtains dataset-macro MRR and Hits@1 of 0.930 and 0.909, compared with 0.926 and 0.903 for Magneto, while reducing mean latency from 123.452 to 15.929 seconds (7.750). Paired analysis finds no statistically significant difference in either MRR or Hits@1. The gate invokes listwise refinement for only 5.665% of source columns and avoids the degradation caused by unconditional refinement. Code and experimental artifacts are available at https://github.com/RazeenLI/JevNexus.

View on arXivView PDF

Code Implementations(1)

RazeenLI/JevNexusOfficial100%
MIT

DeMa — Decision-based Schema Matching

10Shell, PythonSep 23, 20262 weeks agoMIT

Similar Papers

Visual Evidence Under Cross-Examination: Evaluating and Controlling Decision-Level Evidence Use in Vision-Language Models

80% match2026

KC-GenRe: A Knowledge-constrained Generative Re-ranking Method Based on Large Language Models for Knowledge Graph Completion

80% match2024

Bayesian Test-Time Inference of Task-Aligned Similarity from Weak Interactive Feedback

79% match2026

Prompt-Matcher: Leveraging Large Models to Reduce Uncertainty in Schema Matching Results

79% match2024

All Languages Matter: Understanding and Mitigating Language Bias in Multilingual RAG

79% match2026

Reasoning with Preference Constraints: A Benchmark for Language Models in Many-to-One Matching Markets

79% match2025

Cite this paper

@article{li2026correspondences,
  title  = {Correspondences as Decisions: JevNexus for Decision-Centric Schema Matching},
  author = {Runze Li and Hanchen Wang and Ying Zhang and Wenjie Zhang},
  year   = {2026},
  eprint = {2610.09487},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2610.09487v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2610.09487.svg)](https://codeofpaper.com/paper/2610.09487)

Discussion