EIVEN: Efficient Implicit Attribute Value Extraction using Multimodal LLM

arXiv:2404.08886v1 Announce Type: cross
Abstract: In e-commerce, accurately extracting product attribute values from multimodal data is crucial for improving user experience and operational efficiency of retailers. However, previous approaches to multimodal attribute value extraction often struggle with implicit attribute values embedded in images or text, rely heavily on extensive labeled data, and can easily confuse similar attribute values. To address these issues, we introduce EIVEN, a data- and parameter-efficient generative framework that pioneers the use of multimodal LLM for implicit …

abstract arxiv commerce cs.ai cs.cl cs.cv cs.ir cs.lg data e-commerce efficiency embedded experience extraction however images improving llm multimodal multimodal data product retailers struggle text type value values