blinoff's picture
Update README.md
97ee3fb
|
raw
history blame
806 Bytes
metadata
languages:
  - ru
multilinguality:
  - monolingual
size_categories:
  - 10K<n<100K
task_categories:
  - text-classification
  - sentiment-analysis
task_ids:
  - sentiment-classification

Dataset Summary

The dataset contains user reviews about medical facilities.

In total it contains 70,597 reviews. The detailed distribution on sentiment scale is:

  • 41,419 positive reviews;
  • 29,178 negative reviews.

Data Fields

Each sample contains the following fields:

  • review_id;
  • category category of medical facility (one of 48);
  • title: review title;
  • content: review text;
  • sentiment: sentiment (positive or negative);
  • source_url.

Python

import pandas as pd
df = pd.read_json('healthcare_facilities_reviews.jsonl', lines=True)
df.sample(5)