IDK-MRC: Unanswerable Questions for Indonesian Machine Reading Comprehension
This addresses a gap in MRC for Indonesian, a medium- to low-resource language, by providing a dataset that includes unanswerable questions to improve model robustness, though it is incremental as it builds on existing methods for dataset creation.
The paper tackles the lack of unanswerable questions in Indonesian machine reading comprehension (MRC) datasets by creating IDK-MRC, a new dataset with over 10K questions that combines automatic and manual generation, resulting in significant performance improvements for Indonesian MRC models, especially on unanswerable questions.
Machine Reading Comprehension (MRC) has become one of the essential tasks in Natural Language Understanding (NLU) as it is often included in several NLU benchmarks (Liang et al., 2020; Wilie et al., 2020). However, most MRC datasets only have answerable question type, overlooking the importance of unanswerable questions. MRC models trained only on answerable questions will select the span that is most likely to be the answer, even when the answer does not actually exist in the given passage (Rajpurkar et al., 2018). This problem especially remains in medium- to low-resource languages like Indonesian. Existing Indonesian MRC datasets (Purwarianti et al., 2007; Clark et al., 2020) are still inadequate because of the small size and limited question types, i.e., they only cover answerable questions. To fill this gap, we build a new Indonesian MRC dataset called I(n)don'tKnow- MRC (IDK-MRC) by combining the automatic and manual unanswerable question generation to minimize the cost of manual dataset construction while maintaining the dataset quality. Combined with the existing answerable questions, IDK-MRC consists of more than 10K questions in total. Our analysis shows that our dataset significantly improves the performance of Indonesian MRC models, showing a large improvement for unanswerable questions.