metadata

license: mit
task_categories:
  - text-retrieval
language:
  - pt
  - es
tags:
  - e-commerce
  - latin-america
size_categories:
  - 1K<n<10K
configs:
  - config_name: default
    data_files:
      - split: pt
        path: test_pt.jsonl
      - split: es
        path: test_es.jsonl

E-FAQ

We present the E-FAQ, a weakly-supervised dataset consisting of a collection of e-commerce frequently asked questions (FAQs). Each entry i of the dataset is the pair (q_{i}, Q_{i}), in which q_{i} is an anchor question sentence and Q_{i} is a set of questions similar to the anchor. All sentences are uttered in brazilian portuguese or spanish.