Enhancing Generative Information Extraction with Two-step Validation: A Product Attribute Use Case
TL;DR
This work proposes a two-step validation method that integrates a PLM block into the generative IE pipeline and thereby leverages LLMs' correction capability, discovering that such a validation task enhances LLM performance, particularly on the extraction of weakly expressed, low-salience entities that appear sparsely throughout the text.
Abstract
The ability of large language models (LLMs) to process and generate text has introduced potential for applications in information extraction (IE). While it's debated whether LLMs outperform smaller fine-tuned models for classification tasks, their strong generalization capability makes them promising for domains with limited labeled data available for fine-tuning. This advantage is particularly relevant for the emerging application of the digital product passport (DPP), where the problem space is broad but domain-specific data remains scarce. Motivated by this use case, we apply generative IE to the product domain, explicitly addressing efficiency, generalizability, and data privacy constraints. We propose a two-step validation method that integrates a PLM block into the generative IE pipeline and thereby leverages LLMs'correction capability. We discover that such a validation task enhances LLM performance, particularly on the extraction of weakly expressed, low-salience entities that appear sparsely throughout the text. For certain entities, the performance of mid-size models can even reach levels comparable to larger models, and the improvement of first-step PLM predictions also enhance the final LLM output. Nevertheless, the effects on the smallest open-source LLMs (e.g., Llama-3.2 3B) is limited. Based on the findings, we develop a demo application for product information extraction that utilizes locally deployed LLMs, targeting further adaptations to real-world DPP use cases.