home..
Reading Wikipedia to Answer Open-Domain Questions
huyi / December 2022
0. Abstract
task
machine reading at scale: tackle open-domain question answering using Wikipedia tackle opendomain question answering using Wikipedia
challenge
- document retrieval (finding the relevant articles)
- machine comprehension of text (identifying the answer spans from those articles)
1. Introduction
why wikipedia instead of knowledge base?
-
KBs are easier for computers to process but too sparsely populated for open-domain question answering.(Miller,2016)
-
Wikipedia contains up-to-date knowledge that humans are interested in. It is designed, however, for humans – not machines – to read.
machine reading at scale (MRS)
不依赖freebase这种图结构
our approach is generic and could be switched to other collections of documents, books, or even daily updated newspapers.