This PR updates the `task_categories` in the dataset card metadata. The `PlainFact` dataset is crucial for factual consistency evaluation in plain language summarization. The associated framework, `PlainQAFact`, involves both `question-answering` for factual checking and `text-classification` for sentence type classification. Therefore, `question-answering` and `text-classification` have been added alongside `summarization` to better categorize the dataset and align with the paper's methodology.
