Sign InOpen Brain
arXivPaperNeeds Review

OntoAligner-Ensemble: Voting-Based Fusion across Heterogeneous Ontology Alignment Techniques

A two-stage ensemble reconciles outputs from string, graph-embedding, and RAG-based ontology aligners. Composition matters: mixed paradigms favor precision, while LLM-only groups more often favor F1.

arXiv · Aug 31, 2026
Open Source Open MarkdownOpen JSON
Source Summary

OntoAligner-Ensemble applies **voting-based fusion** and then post-fusion selection to candidate correspondences from string, KGE, and RAG-based aligners. It was evaluated on **eight tasks across five OAEI tracks**.

Practical Implication

For pipelines that reconcile structured data, treat ensemble composition as a tunable policy. Cross-paradigm groups generally favored precision, while **homogeneous LLM ensembles** more often delivered higher overall F1.

Agent-Ready Context
OntoAligner-Ensemble applies **voting-based fusion** and then post-fusion selection to candidate correspondences from string, KGE, and RAG-based aligners. It was evaluated on **eight tasks across five OAEI tracks**.

For pipelines that reconcile structured data, treat ensemble composition as a tunable policy. Cross-paradigm groups generally favored precision, while **homogeneous LLM ensembles** more often delivered higher overall F1.

The material provides no per-task scores, costs, or latency results. The findings are specific to ontology alignment, so their value for general coding-agent orchestration remains untested.
Connected Context · Feed7 Judgment

This narrows multi-agent ensemble claims to a specific fusion policy and domain: composition and post-fusion selection change the precision–F1 tradeoff, and heterogeneous ensembles are not automatically superior. Against the prior candidates, it supports typed, tunable aggregation rather than unconstrained agent debate, while leaving general orchestration value, coherence, cost, and latency unresolved.

Split the Labor: Separating Evidence Interpretation from Decision AggregationBoth treat aggregation as an explicit policy, but this candidate warns that naive voting thresholds can drift as evidence grows and proposes calibrated pooling instead.Why We Killed Our Multi-Agent Pipeline — Subbiah Sethuraman and Abhilash Asokan, ZS AssociatesThe failed fixed specialist chain limits extrapolation from successful ontology fusion: bounded candidate outputs can be aggregated, whereas sequential specialists may lose context and coherence.FormalAnalyticGeo: A Neural-Symbolic Based Framework for Multimodal Analytic Geometry Problem GenerationFormalAnalyticGeo reinforces the implementation pattern of compiling heterogeneous agent outputs into a structured representation and applying deterministic checks before accepting results.
Context Map
agentdataresearch#harness-engineering#multi-agent
Uncertainty
The material provides no per-task scores, costs, or latency results. The findings are specific to ontology alignment, so their value for general coding-agent orchestration remains untested.