search

ninglab / eCeLLM

Creative Commons Attribution 4.0 International
31 stars 3 forks source link

dataset #9

Closed dddddjcole closed 3 months ago

dddddjcole commented 4 months ago

Hello, is there a corresponding Chinese data set

BoPeng112 commented 4 months ago

Hi, Thank you very much for your interest. Currently, eCeLLM does not have a Chinese version, but a multilingual eCeLLM will be an exciting next move. We will keep you posted.

dddddjcole commented 4 months ago

Hello! I have some questions about datasets. Can you help me with them?

What does 'Product_Substitute_Identification' mean? for example: query : "pack of hotwheels" product : "Just Play Hot Wheels Ready to Race Car Builder, 29 Pieces" label:No

 query : "diamond earrings for men" 
 product : "Dangle cross earrings in Sterling Silver with CZ diamonds - Handmade in the USA" 
 label: No

Under what standards will it be considered correct?

xinyiling commented 4 months ago

We conduct the product substitute identification task following the paper Shopping Queries Dataset: A Large-Scale ESCI Benchmark for Improving Product Search, which breaks down the product-query relevance into the four classes (Exact, Substitute, Complement, and Irrelevant). As defined in this paper, the product substitute identification task requires the model to identify if the product is somewhat relevant, i.e., it fails to fulfill some aspects of the query but the product can be used as a functional substitute (e.g., fleece for a “sweater” query).