add explanations of non-factual instances

README.md

@@ -9,13 +9,17 @@ task_categories:
 ---
 
 PlainFact-summary is a high-quality human-annotated dataset designed for Plain Language Summarization tasks, along with [PlainQAFact](https://github.com/zhiwenyou103/PlainQAFact) factuality evaluation framework, as described in [PlainQAFact: Automatic Factuality Evaluation Metric for Biomedical Plain Language Summaries Generation](https://huggingface.co/papers/2503.08890). It is collected from the [Cochrane database](https://www.cochranelibrary.com/) sampled from CELLS dataset ([Guo et al., 2024](https://doi.org/10.1016/j.jbi.2023.104580)). 
+In addition to using all factual plain language summaries, we also generate contrasting non-factual examples for each plain language summary. These contrasting examples are perturbed using GPT-4o, following the perturbation criteria for faithfulness introduced in APPLS ([Guo et al., 2024](https://aclanthology.org/2024.emnlp-main.519/)).
 
 We also provided a sentence-level version [PlainFact](https://huggingface.co/datasets/uzw/PlainFact) that split the summaries into sentences with fine-grained explanation annotations. In total, we have 200 plain language summary-abstract pairs.
 
 Here are explanations for the headings:
-- **Target_Sentence**: The plain language sentence/summary.
+- **Factual**: "yes": the plain language summary is factual; "no": the plain language summary is non-factual after applying faithfulness perturbation.
+- **Target_Sentence**: The plain language summary.
 - **Original_Abstract**: The scientific abstract corresponding to each sentence/summary.
 
+> Note: the number of factual and non-factual plain language summaries is the same (200 for each).
+
 You can load our dataset as follows:
 ```python
 from datasets import load_dataset