---
dataset_info:
  features:
  - name: headline
    dtype: string
  - name: text
    dtype: string
  - name: relevance
    dtype: string
  - name: label
    dtype: int32
  splits:
  - name: train
    num_bytes: 8313267.6409710925
    num_examples: 5593
  - name: test
    num_bytes: 3564315.3590289075
    num_examples: 2398
  download_size: 7408311
  dataset_size: 11877583
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
license: apache-2.0
task_categories:
- text-classification
language:
- en
tags:
- economics
- finance
pretty_name: US economy relevance
size_categories:
- 1K<n<10K
---

## Description

This dataset is the HuggingFace version of [this dataset](https://www.kaggle.com/datasets/heeraldedhia/us-economic-news-articles).

It has been reduced to the task of relevance detection. Each text has a label representing whether it is related to the US Economy or not. Hence, it is exclusively focused on the US economy and not the topic "economics" as a global one.

Full credits to the author on Kaggle.