FrameNet Semantic Role Classification by Analogy

2026-03-20

AI Summary

This work addresses the reliance on explicit semantic role labels in FrameNet-based semantic role classification by proposing an efficient modeling approach that operates without any semantic role annotations. The method reformulates the task as a binary classification problem grounded in analogical relations, constructing analogy pairs between lexical units and frame elements. During training, it entirely dispenses with semantic role information, and at inference time, it recovers role labels through a combination of random sampling and analogical transfer. The core innovation lies in formally introducing analogical relations as the foundational mechanism for semantic role classification, complemented by a lightweight neural network architecture and probabilistic distribution inference. The resulting model is structurally simple, converges rapidly, and achieves state-of-the-art performance while maintaining computational efficiency.

Technology Category

• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification

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
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

In this paper, we adopt a relational view of analogies applied to Semantic Role Classification in FrameNet. We define analogies as formal relations over the Cartesian product of frame evoking lexical units (LUs) and frame element (FEs) pairs, which we use to construct a new dataset. Each element of this binary relation is labelled as a valid analogical instance if the frame elements share the same semantic role, or as invalid otherwise. This formulation allows us to transform Semantic Role Classification into binary classification and train a lightweight Artificial Neural Network (ANN) that exhibits rapid convergence with minimal parameters. Unconventionally, no Semantic Role information is introduced to the neural network during training. We recover semantic roles during inference by computing probability distributions over candidates of all semantic roles within a given frame through random sampling and analogical transfer. This approach allows us to surpass previous state-of-the-art results while maintaining computational efficiency and frugality.

Problem

• Semantic Role Classification
• FrameNet
• Analogy
• Artificial Neural Network
• Binary Classification

Innovation

• analogy-based learning
• semantic role classification
• FrameNet
• lightweight neural network
• analogical transfer