Introduction

This project is based on the guidance for EIF matching on aws. We replicated the original pipeline provided by AWS and explored the impact of the language and cultural barriers on the performance of EIF matching.

Datasets

Benchmark

1. Model Results [folder]
   • Amazon
   • Amazon (Translated to Korean)
   • Coupang
2. Annotations [folder]
   • Amazon
   • Coupang
3. Evaluation [folder]
   • Amazon
   • Amazon (Translated to Korean)
   • Coupang

Input data

1. Product Names [folder]
   • Original
     • Amazon Products Dataset [data] [source]
     • Coupang Products Dataset [data]
   • Translated
     • Amazon Products Dataset (Korean) [data]
2. HS Code Dataset [folder]
   • Original
     • Global HS Code Dataset [data] [source]
     • Korean HS Code Dataset [data] [source]
   • Translated
     • Global HS Code Dataset (Korean) [data]

License

This library is licensed under the Apache-2.0 License. See the LICENSE file.