MahaSQuAD: Bridging Linguistic Divides in Marathi Question-Answering
This addresses the problem of limited accessibility to question-answering systems for speakers of low-resource languages like Marathi, though it is incremental as it builds on existing SQuAD data.
The researchers tackled the lack of question-answering datasets for low-resource languages by creating MahaSQuAD, the first full SQuAD dataset for Marathi, with 118,516 training samples and a manually verified gold test set of 500 examples, and developed a generic method for translating SQuAD into any low-resource language.
Question-answering systems have revolutionized information retrieval, but linguistic and cultural boundaries limit their widespread accessibility. This research endeavors to bridge the gap of the absence of efficient QnA datasets in low-resource languages by translating the English Question Answering Dataset (SQuAD) using a robust data curation approach. We introduce MahaSQuAD, the first-ever full SQuAD dataset for the Indic language Marathi, consisting of 118,516 training, 11,873 validation, and 11,803 test samples. We also present a gold test set of manually verified 500 examples. Challenges in maintaining context and handling linguistic nuances are addressed, ensuring accurate translations. Moreover, as a QnA dataset cannot be simply converted into any low-resource language using translation, we need a robust method to map the answer translation to its span in the translated passage. Hence, to address this challenge, we also present a generic approach for translating SQuAD into any low-resource language. Thus, we offer a scalable approach to bridge linguistic and cultural gaps present in low-resource languages, in the realm of question-answering systems. The datasets and models are shared publicly at https://github.com/l3cube-pune/MarathiNLP .