README.md

metadata

license: mit
task_categories:
  - token-classification
task_ids:
  - named-entity-recognition
dataset_info:
  features:
    - name: tokens
      list: string
    - name: tags
      list: string
  splits:
    - name: train
      num_bytes: 1761536
      num_examples: 3361
    - name: validation
      num_bytes: 900871
      num_examples: 1819
  download_size: 472719
  dataset_size: 2662407
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*

---

Proposed Active Learning Data split

Overview

This dataset release represents a proposed and experimental data split(processed in BIO format) designed specifically to support and validate planned active learning (AL) cycles for biomedical Named Entity Recognition (NER).

The current version is not a final benchmark split. Instead, it serves as an initial, controlled setup for testing active learning strategies, model uncertainty sampling, and iterative annotation workflows prior to large-scale development.

Both splits (train and validation) have been carefully curated to ensure coverage of all three target entity types:

• CellLine

• CellType

• Tissue

For details on how the splits were created, please refer to raw data and documentation available here

This BIO format has been generated using a Biomedical transformer-based tokenizer for consistency with downstream model training.

Intended Use

Primary Use

• Supervised NER training for biomedical NLP tasks

Not Intended For

• Clinical or patient-level decision making