A Survey on Table Question Answering: Recent Advances

AbstractTable Question Answering (Table QA) refers to providing precise answers from tables answer a user’s question. In recent years, there have been lot of works on table QA, but is lack comprehensive surveys this research topic. Hence, we aim provide an overview available datasets and representative methods in QA. We classify existing for QA into five categories according their techniques, which include semantic-parsing-based, generative, extractive, matching-based, retriever-reader-based methods. Moreover, because still challenging task methods, also identify outline several key challenges discuss the potential future directions QA.KeywordsNatural language processingTable QASemantic parsing