机器阅读理解研究与进展

RESEARCH AND PROGRESS ON MACHINE READING COMPREHENSION

针对目前机器阅读理解的研究进展,对机器阅读理解的研究背景和国内外研究现状进行详细介绍,着重介绍国内外主流的大规模机器阅读理解数据集,以及在各个数据集上的评价指标。介绍神经机器阅读理解模型,并对向量化、编码、注意力机制、答案预测模块做了详细的介绍。总结当前机器阅读理解所面临的问题,并展望未来的发展趋势。

In view of the current research progress of machine reading comprehension,the paper introduced the research background of machine reading comprehension and the research status at home and abroad in detail.It focused on the mainstream large-scale machine reading comprehension datasets at home and abroad,as well as the evaluation indicators on each dataset.The neural machine reading comprehension model was introduced,including the vectorization,coding,attention mechanism and answer prediction module.The paper summarized the current problems and looked forward to the future development trend.