Child Protection Helpline Case Summarization Dataset
Overview
This dataset provides 1,000 synthetic child protection helpline call transcripts paired with concise, structured summaries for training sequence-to-sequence summarization models. It simulates real-world calls reporting various forms of abuse and exploitation in East African contexts while ensuring no real cases or personal data are included
The data was created to fine-tune FLAN-T5 and related models for rapid… See the full description on the dataset page: https://huggingface.co/datasets/openchs/sum-synthetic-data-v1.