Downstream Transformer Generation of Question-Answer Pairs with Preprocessing and Postprocessing Pipelines
This addresses the need for automated question generation in educational or content analysis domains, but it is incremental as it builds on existing transformer methods with added pipelines.
The paper tackles the problem of generating question-answer pairs from articles by introducing TP3, a system that finetunes pretrained transformers and uses preprocessing and postprocessing pipelines, resulting in satisfactory numbers of high-quality QAPs on the Gaokao-EN dataset.
We present a system called TP3 to perform a downstream task of transformers on generating question-answer pairs (QAPs) from a given article. TP3 first finetunes pretrained transformers on QAP datasets, then uses a preprocessing pipeline to select appropriate answers, feeds the relevant sentences and the answer to the finetuned transformer to generate candidate QAPs, and finally uses a postprocessing pipeline to filter inadequate QAPs. In particular, using pretrained T5 models as transformers and the SQuAD dataset as the finetruning dataset, we show that TP3 generates satisfactory number of QAPs with high qualities on the Gaokao-EN dataset.