Layout-Aware Parsing Meets Efficient LLMs: A Unified, Scalable Framework for Resume Information Extraction and Evaluation

2025-10-10

AI Summary

Resume information extraction faces three key challenges: high layout heterogeneity, prohibitive inference costs of large language models (LLMs), and the absence of a unified, reproducible evaluation standard. To address these, we propose a two-stage framework integrating layout-aware parsing with a lightweight LLM. In Stage I, a fine-tuned layout parser performs fine-grained structural understanding of resume documents. In Stage II, an instruction-tuned 0.6B-parameter LLM—enhanced with parallel prompting—executes efficient information extraction and automated evaluation. We further introduce a novel, unified, and reproducible fine-grained evaluation benchmark. Experiments demonstrate that our method matches the accuracy of state-of-the-art large models (e.g., Llama-3-70B) while reducing inference latency by 92% and FLOPs by 98%. The system has been deployed at scale on Alibaba’s Intelligent HR Platform, supporting real-time applications across multiple business lines.

Technology Category

• Natural Language Processing: Information Extraction
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Automated resume information extraction is critical for scaling talent acquisition, yet its real-world deployment faces three major challenges: the extreme heterogeneity of resume layouts and content, the high cost and latency of large language models (LLMs), and the lack of standardized datasets and evaluation tools. In this work, we present a layout-aware and efficiency-optimized framework for automated extraction and evaluation that addresses all three challenges. Our system combines a fine-tuned layout parser to normalize diverse document formats, an inference-efficient LLM extractor based on parallel prompting and instruction tuning, and a robust two-stage automated evaluation framework supported by new benchmark datasets. Extensive experiments show that our framework significantly outperforms strong baselines in both accuracy and efficiency. In particular, we demonstrate that a fine-tuned compact 0.6B LLM achieves top-tier accuracy while significantly reducing inference latency and computational cost. The system is fully deployed in Alibaba's intelligent HR platform, supporting real-time applications across its business units.

Problem

• Addresses extreme heterogeneity in resume layouts and content
• Reduces high cost and latency of large language models
• Provides standardized datasets and evaluation tools for extraction

Innovation

• Layout-aware parser normalizes diverse resume formats
• Efficient LLM uses parallel prompting and instruction tuning
• Two-stage automated evaluation framework with benchmark datasets