Reading Comprehension in Czech via Machine Translation and Cross-lingual Transfer
This enables reading comprehension in Czech and potentially other languages without costly manual annotation, though it is incremental as it builds on existing cross-lingual transfer methods.
The paper tackled building reading comprehension systems for Czech without manually annotated Czech training data by automatically translating English datasets and evaluating cross-lingual transfer models. The result showed that a model trained on English data and evaluated on Czech achieved performance only about 2 percentage points worse than one trained on translated Czech data.
Reading comprehension is a well studied task, with huge training datasets in English. This work focuses on building reading comprehension systems for Czech, without requiring any manually annotated Czech training data. First of all, we automatically translated SQuAD 1.1 and SQuAD 2.0 datasets to Czech to create training and development data, which we release at http://hdl.handle.net/11234/1-3249. We then trained and evaluated several BERT and XLM-RoBERTa baseline models. However, our main focus lies in cross-lingual transfer models. We report that a XLM-RoBERTa model trained on English data and evaluated on Czech achieves very competitive performance, only approximately 2 percent points worse than a~model trained on the translated Czech data. This result is extremely good, considering the fact that the model has not seen any Czech data during training. The cross-lingual transfer approach is very flexible and provides a reading comprehension in any language, for which we have enough monolingual raw texts.