A framework for merging and ranking of answers in DeepQA

From Brede Wiki
Jump to: navigation, search
Paper (help)
A framework for merging and ranking of answers in DeepQA
Authors: David C. Gondek, Adam Lally, Aditya Kalyanpur, J. William Murdock, Pablo A. Duboue, Lei Zhang, Yue Pan, Zhao Ming Qiu, Chris Welty
Citation: IBM Journal of Research and Development 56 (3/4): 14. 2012 May/July
Database(s): Google Scholar cites
DOI: 10.1147/JRD.2012.2188760.
Link(s): http://www.andrew.cmu.edu/user/ooo/watson/14%20a%20framework%20for%20merging%20and%20ranking%20answers.pdf
Search
Web: Bing Google Yahoo!Google PDF
Article: BASE Google Scholar PubMed
Restricted: DTU Digital Library
Other: NIF
Services
Format: BibTeX
Extract: Talairach coordinates from linked PDF: CSV-formated wiki-formated

A framework for merging and ranking of answers in DeepQA describes a stage in the IBM Watson question answering system where candidate answers are ranked.

The researchers experimented with a number of different machine learning techniques, but settled on regularized logistic regression which consistently performed better.

[edit] Related papers

  1. Introduction to "This is Watson"
Personal tools