DivLog proposes a log parsing framework based on large language models (LLMs) and in-context learning (ICL). It samples diverse logs offline, selects appropriate examples for each target log during parsing, and generates log templates without model tuning. DivLog achieves state-of-the-art performance with high accuracy metrics across 16 datasets. The framework enhances the quality of generated log templates and demonstrates stability and robustness in log analysis tasks.
A otro idioma
del contenido fuente
arxiv.org
Ideas clave extraídas de
by Junjielong X... a las arxiv.org 03-01-2024
https://arxiv.org/pdf/2307.09950.pdfConsultas más profundas