Galaxy_clients / README.md
ekrombouts's picture
Upload dataset
d020d43 verified
metadata
language:
  - nl
license: cc-by-sa-4.0
size_categories:
  - 10K<n<100K
task_categories:
  - text-generation
pretty_name: Clientprofiles of Nursing Home Residents
dataset_info:
  features:
    - name: ct_id
      dtype: int64
    - name: ward
      dtype: string
    - name: name
      dtype: string
    - name: dementia_type
      dtype: string
    - name: physical
      dtype: string
    - name: adl
      dtype: string
    - name: mobility
      dtype: string
    - name: behavior
      dtype: string
    - name: complications
      dtype: string
    - name: num_months
      dtype: int64
  splits:
    - name: train
      num_bytes: 30981
      num_examples: 96
  download_size: 19462
  dataset_size: 30981
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
tags:
  - medical

Galaxy Datasets

The Galaxy Datasets are a collection of four synthetic datasets created for NLP experiments, featuring client records in a nursing home setting. Each dataset serves a specific purpose and builds upon the previous one, providing a useful resource for various NLP tasks.

Galaxy_clients contains profiles of nursing home residents, including additional information such as the number of months before data generation for each client and the complications they experienced during their stay. This dataset provides the foundational client information necessary for generating realistic scenarios and care reports.

Galaxy_scenarios uses the information from Galaxy_clients to create detailed scenarios for each client. These scenarios describe potential situations and complications, serving as the basis for generating synthetic care notes in the next dataset.

Galaxy_records consists of client records with synthetic care notes based on the scenarios outlined in Galaxy_scenarios. These care notes provide detailed and structured documentation of the care each client received, simulating real-world healthcare documentation for NLP model training and testing.

Galaxy_summaries contains monthly summaries of the notes in Galaxy_records. Initially created to support the generation process of Galaxy_records, these summaries can also be used to train open-source models in text summarization, offering a valuable resource for developing and refining summarization techniques.

For more information and access to the datasets, visit the GitHub page.