Enhancing E-Commerce through Transformer-Based Large Language Models: Automating Multilingual Product Descriptions for Improved Customer Engagement

This study looks at how transformer-based Large Language Models (LLMs) are used in e-commerce—a field where online shopping is becoming more and more prevalent—to automate the creation of product descriptions. This study looks at how the online companies translate product data that is organised internally into semi-structured storefront presentations that are visually appealing. This study investigates the possibility that these LLMs could properly identify pertinent product features and produce coherent descriptions by using models like BigScience Large Open-science Open-access Multilingual Language Model (BLOOM)—which are renowned for their advanced language processing capabilities. The study also explores the difficulties associated with localization, offering suggestions for modifying these models to provide product descriptions in many languages, increasing their usefulness in international marketplaces. This study functions as a proof of concept, showcasing the viability and efficacy of LLMs in enhancing customer engagement through improved product descriptions despite the complexity of the data and the usage of a limited model. The results demonstrate how LLMs can close the gap between product data and customer engagement—making a substantial theoretical and practical contribution to e-commerce.