Saliency Map-Guided Knowledge Discovery for Subclass Identification with LLM-Based Symbolic Approximations

2025-11-10

AI Summary

To address the challenge of identifying latent subclasses in multiclass time-series classification, this paper proposes a saliency-map-guided neuro-symbolic fusion framework. First, the original multiclass task is decomposed into label-entailment-driven binary subproblems; discriminative temporal segments are localized via neural network gradient-based saliency maps. Second, these salient regions guide multi-configuration clustering and cluster-center extraction. Finally, large language models (LLMs) perform symbolic approximation and fuzzy knowledge graph matching to generate interpretable semantic descriptions of discovered subclasses. This work is the first to incorporate gradient saliency information into subclass discovery, achieving a principled balance between signal modeling fidelity and symbolic reasoning interpretability. Extensive experiments on multiple benchmark datasets demonstrate that our method significantly outperforms purely signal-driven baselines in both subclass identification accuracy and semantic plausibility.

Technology Category

• Machine Learning: Neuro-Symbolic Learning
• Computer Vision: Visual Reasoning & Symbolic Representations
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

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

Abstract

This paper proposes a novel neuro-symbolic approach for sensor signal-based knowledge discovery, focusing on identifying latent subclasses in time series classification tasks. The approach leverages gradient-based saliency maps derived from trained neural networks to guide the discovery process. Multiclass time series classification problems are transformed into binary classification problems through label subsumption, and classifiers are trained for each of these to yield saliency maps. The input signals, grouped by predicted class, are clustered under three distinct configurations. The centroids of the final set of clusters are provided as input to an LLM for symbolic approximation and fuzzy knowledge graph matching to discover the underlying subclasses of the original multiclass problem. Experimental results on well-established time series classification datasets demonstrate the effectiveness of our saliency map-driven method for knowledge discovery, outperforming signal-only baselines in both clustering and subclass identification.

Problem

• Identifying latent subclasses in time series classification tasks
• Transforming multiclass classification into binary problems via label subsumption
• Discovering underlying subclasses using saliency maps and symbolic approximations

Innovation

• Saliency maps guide subclass discovery process
• Transform multiclass into binary classification problems
• LLM performs symbolic approximation for knowledge graphs