---
title: "Bridging Structure and Semantics: Knowledge Graph Verification via Triple-to-Text Alignment"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.07780v1"
permalink: "https://feed.astralor.com/items/itm_b7e260d88b437b51d2fb5ea0"
published: "2026-11-05T10:19:27.000Z"
score: 56
channel: "ai"
tags: ["安全","Bridging Structure","Semantics","Knowledge Graph Verification","Triple-to-Text Alignment Semantic","Graph-based","Semantic","PLMs"]
---

# Bridging Structure and Semantics: Knowledge Graph Verification via Triple-to-Text Alignment

## At a glance

Semantic accuracy is an important dimension of knowledge graph quality. Graph-based scoring methods exploit relational structure, while textual representations allow classifiers to draw on knowledge encoded by pre-train…

## Key points

- Semantic accuracy is an important dimension of knowledge graph quality.
- Graph-based scoring methods exploit relational structure, while textual representations allow classifiers to draw on knowledge encoded by pre-trained language models (PLMs).
- We present GenVer, a generative verifier that scores triples using both their serialized form and generated natural language representations (NLRs).

## Full text

Semantic accuracy is an important dimension of knowledge graph quality. Graph-based scoring methods exploit relational structure, while textual representations allow classifiers to draw on knowledge encoded by pre-trained language models (PLMs). We present GenVer, a generative verifier that scores triples using both their serialized form and generated natural language representations (NLRs). A triple-text alignment procedure selects training pairs for a sequence-to-sequence generator; a BERT-based classifier then combines features from each triple and its generated NLR. On four FactFB variants, constructed from FB15k using perturbed negative triples, GenVer obtains the highest classification accuracy among the evaluated baselines, ranging from 88.56% to 91.52%. The experiments suggest that generated representations can help triple classification, while the case analysis also exposes semantic drift and high-confidence errors.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_b7e260d88b437b51d2fb5ea0