Leveraging the Power of Large Language Models in Entity Linking via Adaptive Routing and Targeted Reasoning

2025-10-22

AI Summary

To address the high inference cost and substantial fine-tuning overhead of few-shot entity linking (EL) methods, this paper proposes ARTER—a novel adaptive routing framework. ARTER dynamically partitions mentions into “easy” and “hard” categories via an adaptive routing mechanism, dispatching them respectively to a lightweight linker (e.g., ReFinED) and a large language model (LLM). It integrates embedding representation, candidate generation, context-aware scoring, and multi-signal classification to enable targeted LLM inference. Crucially, ARTER avoids full-model fine-tuning and generic LLM invocation. Evaluated on six standard benchmarks, it achieves an average F1-score improvement of 2.53% (up to +4.47%) on five datasets, doubles inference speed over end-to-end LLM baselines, and significantly reduces token consumption—thereby balancing efficiency, accuracy, and practical deployability.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Entity Linking (EL) has traditionally relied on large annotated datasets and extensive model fine-tuning. While recent few-shot methods leverage large language models (LLMs) through prompting to reduce training requirements, they often suffer from inefficiencies due to expensive LLM-based reasoning. ARTER (Adaptive Routing and Targeted Entity Reasoning) presents a structured pipeline that achieves high performance without deep fine-tuning by strategically combining candidate generation, context-based scoring, adaptive routing, and selective reasoning. ARTER computes a small set of complementary signals(both embedding and LLM-based) over the retrieved candidates to categorize contextual mentions into easy and hard cases. The cases are then handled by a low-computational entity linker (e.g. ReFinED) and more expensive targeted LLM-based reasoning respectively. On standard benchmarks, ARTER outperforms ReFinED by up to +4.47%, with an average gain of +2.53% on 5 out of 6 datasets, and performs comparably to pipelines using LLM-based reasoning for all mentions, while being as twice as efficient in terms of the number of LLM tokens.

Problem

• Reducing expensive LLM reasoning in entity linking
• Minimizing training dependency through adaptive routing
• Balancing accuracy and efficiency in few-shot entity linking

Innovation

• Adaptive routing categorizes mentions into easy and hard cases
• Targeted LLM reasoning handles only the difficult entity cases
• Hybrid pipeline combines embedding and LLM-based complementary signals