Question Anwsering In General

Posted by Kaiyuan Chen on October 23, 2017



query: keywords that system should look at and search for


Anwser type classification: determine answer type. Word in a question can give extra information: question headword & anwser type word.

Query formulation: rephrase the question: rephrase the question that look like substring of declarative answers

passage retrieval: extract potential answer of the passage passages are ranked based on named entities, keywords proximity

answer processing: pattern extraction

Knowledge based QA

have a databased relations e.g. infoboxes that has structued tables associated with wiki’s articles

The supervised approach can be extended to deal with more complex questions that are not just about single relations. argmax(lx:state(x) ^ borders(x;texas);lx:size(x))

most methods make some use of web text, either via semi-supervised methods like distant supervision or unsupervised methods like open information extraction, both introduced in Chapter 20