PDF | IBM Research undertook a challenge to build a computer system that could compete at the human champion level in real time on the American TV quiz. Build watson: An overview of DeepQA for the Jeopardy! The DeepQA project ( ) is aimed at illustrating how the advancement and. @article{journals/aim/FerrucciBCFGKLMNPSW10, added-at = {T +}, author = {Ferrucci, David A. and Brown, Eric W. and.

Author: Kigarisar Doum
Country: Syria
Language: English (Spanish)
Genre: Travel
Published (Last): 12 April 2008
Pages: 243
PDF File Size: 2.33 Mb
ePub File Size: 5.30 Mb
ISBN: 925-7-63837-953-4
Downloads: 23189
Price: Free* [*Free Regsitration Required]
Uploader: Mooguran

Percent answered whether or not Watson can win one or two games is the percentage of questions it chooses to answer against top-ranked humans in real time.

Virginia and Indiana explicit LATs in the 20, question sample. Watson Research Center Aditya A.

This is roughly between 1 and 6 seconds ly a laboratory exercise. One of the goals of the sys- evidence and produce a score that ovdrview to tem design, therefore, is to tolerate noise wqtson the how well evidence supports a candidate answer for early stages of the pipeline and drive up precision a given question. QA process, from focus and LAT determination, to We refer to search performed in hypothesis gen- passage and answer scoring.

Jeopardy clues are straightforward assertional forms of questions. Film of a typical day in wwtson life of the Bea- Answer: The seepqa of the challenge includes fielding a real-time automatic contestant on the show, not merely a laboratory exercise.

Evidence Profiles overvieew Two Candidate Answers. The Deep- identify answer types for a question, and candidate QA system at the time had accuracy above 50 per- answer-generation components that identify cent on Jeopardy.

For example, a ferent types of sources including unstructured deeepqa, lightweight scorer may compute the likelihood of semistructured text, and triple stores.


While potentially compelling for a pub- Figure 2 shows a plot of precision versus percent lic contest, a small number of games does not rep- attempted curves for two theoretical systems. The system generates the correct answer as a candidate answer may generate a number of candidate answer vari- for 85 percent of the questions somewhere within ants from the same title based on substring analy- the top ranked candidates.

The questions each question as possible interpretations. Diplomatic Relations questions and Special Instructions questions. In both phases sets of scores shows why. Chile shares its Cervantes, the correct answer, was born in longest land border with this country.

Building Watson: An Overview of the DeepQA Project | Nico Schlaefer –

The four countries in the world that mine a correct answer. The author s agree that if anyone brings any claim or action alleging facts that, if true, constitute a breach of any of byilding foregoing warranties, the author s will hold harmless and indemnify AAAI, their grantees, their licensees, and their distributors against any liability, whether under judgment, decree, or compromise, and any legal fees and expenses wztson out of that claim or actions, and the undersigned will cooperate fully in any defense AAAI may make to such claim or action.

Light or Photons answers must rhyme with one another.

In to serve as training data for learning techniques. While we believe the Jeopardy Challenge leading, for a computer.

The Both present very interesting challenges from an most frequent explicit LATs cover less than 50 AI perspective but were put out of scope for this percent of the data. Natural Langage Engi- described in this paper. Special instruction questions are those that are answer must be inferred by the context. Tge archaic term deelqa a projdct or annoy- Subclue 1: Secretary Chase just submitted this to me for For example: The soft obscurityits aliases, and so on.


Even if the question tion that, if replaced by the answer, makes the ques- did not need to be decomposed to determine an tion a stand-alone statement.

Building Watson: An Overview of the DeepQA Project

At the end of the 30 be in the form of a question. Scoring algo- stage as a candidate, the system has no hope of rithms determine the degree of certainty ubilding answering the question. Lexical Answer Type Frequency. Dredze, This is particularly confusing to ranking tech- Crammer, and Pereira Leveraging category information is playing chess.

They must answer the question, but the response must have 30 seconds to respond. Confi- openly advance QA research.

The and had a week to produce results for ques- search-based system has better performance at tions. A requirement of the Jeopardy formance of a system based purely on text search, Challenge is that th system be self-contained and using terms in the question as queries and search does not link to live web search.

He is a lead devel- Open-Domain Question-Answering. It Approach to Unstructured Information Processing in the is this team who are responsible for the work Corporate Research Environment.

Perfect confidence estimation upper line and no buildinb estimation lower line. The scoring step is where the bulk of the score. A Large-Scale Investment in 5. A lytics to evaluate the supporting evidence. New Directions in Question- assigns to short queries, which typically are not suffi- Answering. Based on our analysis of those games, he precision was 13 percent.