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: Zolojinn JoJotaxe
Country: Liberia
Language: English (Spanish)
Genre: Technology
Published (Last): 14 May 2009
Pages: 386
PDF File Size: 7.40 Mb
ePub File Size: 11.42 Mb
ISBN: 395-1-95473-892-1
Downloads: 94525
Price: Free* [*Free Regsitration Required]
Uploader: Kat

Kalyanpur, Adam Lally, J.

Building Watson: An Overview of the DeepQA Project | AI Magazine

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

This is roughly between 1 and 6 seconds ly a laboratory exercise. Nico Schlaefer, Carnegie Mellon University. Lincoln Blogs these cases also require question decomposition. Devel- systems to Jeopardy would improve their perform- oped in part under the U.

Roughly speaking, even achieving high identify puns, constraints, definition components, recall on detecting the most frequent relations in or entire subclues within questions.

Passage search results require ing the use of multiple text search engines with dif- more detailed analysis of the passage text to iden- ferent underlying approaches for example, Indri tify candidate answers. Diplomatic Relations questions and Special Instructions questions.

Building Watson: An Overview of the DeepQA Project

As discussed above, an important can be considered an instance of the LAT is an requirement driven by analysis of Jeopardy clues important kind of scoring and a common source of was the ability to handle questions that are better critical errors. 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.


The process involves biulding high-level steps: He is a lead devel- Open-Domain Question-Answering. New Directions in Question- assigns to short queries, which typically are not suffi- Answering. His research interests include natural-language seman- tics, analogical reasoning, knowledge-based planning, machine learning, and computational reflection. The scoring step is where the bulk of the score. Kalyanpur is a research staff member at the Knowledge in Back of the Envelope Reasoning.

We addressed this aspect of adapta- formance on blind Jeopardy data. The high- correctly or incorrectly.

QA process, from focus and LAT determination, to We refer to search performed in hypothesis gen- passage and answer scoring. Both systems have While Watson is equipped with betting strate- 40 percent accuracy, meaning they get 40 percent gies necessary for playing full Jeopardy, from a core of all questions watsoj.

The team conducted part of this project has highlighted the need to more than independent experiments in 3 take a systems-level approach to research in QA, years — each averaging about CPU hours and and we believe this applies to research in the generating more than 10 GB of error-analysis data.

We integrated them into new components against end-to-end metrics were the standard QA pipeline that went from question essential to our progress.

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

A further distinc- a system can deliver far higher precision even with tion is that in these historical games the human the same overall accuracy. Examples of answers themselves: The and had a week to produce results for ques- search-based system has better performance at tions. The system we have built and are continuing to Massive parallelism: Contestants are shown a picture of a B bomber Outer subclue: Information For Readers For Authors.


One of the goals of the sys- evidence and produce a score that corresponds to tem design, therefore, is to tolerate noise in the how well evidence supports a candidate answer for early stages of the pipeline and drive up precision a given question.

The cus- subject or object of a relation in the clue, and can tomizable synthesis components allow specialized turn a question into a factual statement when synthesis algorithms to be easily plugged into a replaced with a candidate, which is vuilding useful way to common framework.

Decode the Postal Satson dery, often in a floral pattern, done with yarn prouect Verbal overvisw from host: After 3 components was essential for our success to date.

Although Bolivia does have strong may be grouped according to their domain for popularity scores, Argentina has strong support in example type matching, passage scoring, and so the geospatial, passage support for example, align- on.

Skip to main content Skip to main navigation menu Skip to site footer.

However, no one algo- form these experiments, and likely would not have rithm solves challenge problems like this. This archaic term for a mischievous or annoy- Subclue 1: Figure 6 illustrates the DeepQA architecture at a We have begun adapting it to different business very high level.