A Boring-yet-effective Approach for the Product Ranking Task of the Amazon KDD Cup 2022

TL;DR

This paper presents a straightforward approach using large language models like mT5 for product ranking in the Amazon KDD Cup 2022, achieving near-top performance with minimal task-specific modifications.

Contribution

The work demonstrates that a simple, minimally adapted large language model approach can achieve competitive results in e-Commerce ranking tasks.

Findings

Best model was less than 0.004 nDCG@20 below the top submission

Top teams achieved nDCG@20 close to 0.90

More challenging datasets are needed for better evaluation

Abstract

In this work we describe our submission to the product ranking task of the Amazon KDD Cup 2022. We rely on a receipt that showed to be effective in previous competitions: we focus our efforts towards efficiently training and deploying large language odels, such as mT5, while reducing to a minimum the number of task-specific adaptations. Despite the simplicity of our approach, our best model was less than 0.004 nDCG@20 below the top submission. As the top 20 teams achieved an nDCG@20 close to .90, we argue that we need more difficult e-Commerce evaluation datasets to discriminate retrieval methods.