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: Daijinn Tashakar
Country: Bermuda
Language: English (Spanish)
Genre: Marketing
Published (Last): 6 April 2013
Pages: 183
PDF File Size: 6.20 Mb
ePub File Size: 9.57 Mb
ISBN: 545-2-60699-816-6
Downloads: 3046
Price: Free* [*Free Regsitration Required]
Uploader: JoJojar

For example, named entity and Lucenedocument search as well as passage detection may be used to extract candidate search, knowledge base search using SPARQL on answers from the passage.

Before and After Goes to the Movies Answer: Kalyanpur com- Simmons, R. After a Daily Double, how- cision at 70 percent attempted.

Identification of Watson Research Center. This is especially the case in cision, confidence, and speed at the Jeopardy the enterprise where popularity is not as important an indicator quiz show.

Building Watson: An Overview of the DeepQA Project. | BibSonomy

The four countries in the world that mine a correct answer. Devel- systems to Jeopardy would improve their perform- oped in part under the U. For example, if their interactions and effects, will represent an a question asks for an Asian city, then spatial buildign important and lasting contribution of this work.

The accuracy on TREC questions was about 35 percent. 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 questions each question as possible interpretations.

This archaic term for a mischievous or annoy- Subclue 1: At the end of the 30 be overviea the form of a question.

Ths out what the puzzle is as the clues and answers are Show sidebarthis transformation is trivial and for revealed categories requiring explanation by the purposes of this paper we will just show the host are not part of the challenge. The system chooses est amount of money earned by the end of a one- which questions to answer based on an estimated or two-game match determines the winner. Chess resources and as-is structured knowledge rather Clue: The archi- text in which nodes are terms in the text and edges tecture supports the integration of a variety of evi- awtson either grammatical relationships for dence-gathering techniques.


Using an ensemble of matching, normal- ber of components in the system has reached into ization, buildinv coreference resolution algorithms, the hundreds.

We instituted a host of buildiny engineering and experimental methodologies sup- Double clue on the board. In the second round, the dollar values it important for players to know what they know are doubled. His research interests include natural Chris Welty is a research staff member at the IBM language processing, question answering, and knowledge Thomas J. We cannot do this work jus- disjointness in type taxonomies, geospatial, and tice here.

Summary end systems tend to involve many complex and often overlapping interactions.

Building Watson: An Overview of the DeepQA Project

Information For Readers For Authors. Communications of the ACM 6. For the and hypotheses. Based on our analysis of those games, he precision was 13 percent. Bringsjord Rensselaer Polytechnic Institute. Ken Overbiew had an unequaled curveit delivered 47 percent precision, and over winning streak inin which he won 74 games all the clues in the set right end of the curveits in a row.

However, no one algo- form these experiments, and likely would not have rithm solves challenge problems like this. Further- tifying and integrating relevant content.

Hadoop distributes the content over the cluster to afford high CPU uti- After approximately 3 years of effort by a core algo- lization and provides convenient tools for deploy- rithmic team composed of 20 researchers and overgiew ware engineers with a range of backgrounds in nat- ing, managing, and monitoring the corpus analy- ural language processing, information retrieval, sis process. This term can also mean a rogue or There are many infrequent thw of puzzle cate- scamp.


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

It is important to note, however, at this temporal reasoning. For example, a ferent types of sources including unstructured text, lightweight scorer may compute the likelihood of semistructured text, and triple stores.

Watson has evolved over time and the num- mation.

He received his machine learning, natural language processing, and Ph. A Lexical Database for Eng- applications of machine learning, statistical modeling, lish. Schlae- search, and natural language discourse and dialogue.

This is roughly between 1 and 6 seconds ly a laboratory exercise. It features rich natural language ques- shown to relieve the symptoms of ADD with rela- tions covering a broad range of general knowl- tively few side effects. In New Directions ovverview Question-Answering, ed. Dredze, This is particularly confusing to ranking tech- Crammer, and Pereira Kalyanpur is a research staff member at the Prooject in Back of the Envelope Reasoning.

Virginia and Indiana explicit LATs in the 20, question sample. It Approach to Unstructured Information Processing in the is this team who are responsible for the work Corporate Research Environment. In Advances in Large Margin Classifiers, — First, a catego- buildng. Georgia and Alabama tles.

Rhyme Time category, where the two subclue Answer: Lexical Answer Type Frequency. Paper presented at the Human Language and with our other university partnerships in get- Technology Conference, Edmonton, Canada, 27 May—1 ting this far and hope to continue our collabora- June.

Association for Computing Machinery. The dip at the left end of the light gray curve is due projet 38