jurafsky&martin_3rdEd_17 (1).pdf

Phrases of a question as part of the process of

Info icon This preview shows pages 412–414. Sign up to view the full content.

View Full Document Right Arrow Icon
phrases of a question as part of the process of finding an answer ( Fader et al. 2013 , Berant and Liang 2014 ). 27.3 Using multiple information sources: IBM’s Watson Of course there is no reason to limit ourselves to just text-based or knowledge-based resources for question answering. The Watson system from IBM that won the Jeop- ardy! challenge in 2011 is an example of a system that relies on a wide variety of resources to answer questions. Document Document Document (1) Question Processing From Text Resources Focus Detection Lexical Answer Type Detection Question Document and Passsage Retrieval passages Document Document Document Question Classification Parsing Named Entity Tagging Relation Extraction Coreference From Structured Data Relation Retrieval DBPedia Freebase (2) Candidate Answer Generation Candidate Answer Candidate Answer Candidate Candidate Candidate Candidate Answer Candidate Candidate Candidate Answer Candidate Answer Candidate Candidate Answer (3) Candidate Answer Scoring Evidence Retrieval and scoring Answer Extraction Document titles Anchor text Text Evidence Sources (4) Confidence Merging and Ranking Text Evidence Sources Time from DBPedia Space from Facebook Answer Type Answer and Confidence Candidate Answer + Confidence Candidate Answer + Confidence Candidate Answer + Candidate Answer Candidate Answer + Confidence Logistic Regression Answer Ranker Merge Equivalent Answers Figure 27.9 The 4 broad stages of Watson QA: (1) Question Processing, (2) Candidate Answer Generation, (3) Candidate Answer Scoring, and (4) Answer Merging and Confidence Scoring. Figure 27.9 shows the 4 stages of the DeepQA system that is the question an- swering component of Watson. The first stage is question processing . The DeepQA system runs parsing, named entity tagging, and relation extraction on the question. Then, like the text-based systems in Section 27.1 , the DeepQA system extracts the focus , the answer type (also called the lexical answer type or LAT ), and performs question classification and question sectioning . Consider these Jeopardy! examples, with a category followed by a question: Poets and Poetry: He was a bank clerk in the Yukon before he published “Songs of a Sourdough” in 1907. THEATRE: A new play based on this Sir Arthur Conan Doyle canine classic opened on the London stage in 2007. The questions are parsed, named entities are extracted ( Sir Arthur Conan Doyle identified as a PERSON , Yukon as a GEOPOLITICAL ENTITY , “Songs of a Sour- dough” as a COMPOSITION ), coreference is run ( he is linked with clerk ) and rela- tions like the following are extracted: authorof(focus,“Songs of a sourdough”) publish (e1, he, “Songs of a sourdough”)
Image of page 412

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
27.3 U SING MULTIPLE INFORMATION SOURCES : IBM’ S W ATSON 413 in (e2, e1, 1907) temporallink(publish(...), 1907) Next DeepQA extracts the question focus , shown in bold in both examples. The focus focus is the part of the question that co-refers with the answer, used for example to align with a supporting passage. The focus is extracted by hand-written rules—made possible by the relatively stylized syntax of Jeopardy! questions—such as a rule
Image of page 413
Image of page 414
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern