Machine Reading Comprehension: The Role of Contextualized Language Models and Beyond

  • 2020-05-13 10:58:50
  • Zhuosheng Zhang, Hai Zhao, Rui Wang
  • 1

Abstract

Machine reading comprehension (MRC) aims to teach machines to read andcomprehend human languages, which is a long-standing goal of natural languageprocessing (NLP). With the burst of deep neural networks and the evolution ofcontextualized language models (CLMs), the research of MRC has experienced twosignificant breakthroughs. MRC and CLM, as a phenomenon, have a great impact onthe NLP community. In this survey, we provide a comprehensive and comparativereview on MRC covering overall research topics about 1) the origin anddevelopment of MRC and CLM, with a particular focus on the role of CLMs; 2) theimpact of MRC and CLM to the NLP community; 3) the definition, datasets, andevaluation of MRC; 4) general MRC architecture and technical methods in theview of two-stage Encoder-Decoder solving architecture from the insights of thecognitive process of humans; 5) previous highlights, emerging topics, and ourempirical analysis, among which we especially focus on what works in differentperiods of MRC researches. We propose a full-view categorization and newtaxonomies on these topics. The primary views we have arrived at are that 1)MRC boosts the progress from language processing to understanding; 2) the rapidimprovement of MRC systems greatly benefits from the development of CLMs; 3)the theme of MRC is gradually moving from shallow text matching to cognitivereasoning.

 

Quick Read (beta)

loading the full paper ...