Coupled Bayesian Sets Algorithm for Semi-supervised...
Transcript of Coupled Bayesian Sets Algorithm for Semi-supervised...
![Page 1: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/1.jpg)
Coupled Bayesian Sets Algorithm for Semi-supervised Learning and
Information Extraction
Saurabh Verma Baranas Hindu University, India
Estevam R. Hruschka Jr.
Federal University of São Carlos, Brazil
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 2: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/2.jpg)
http://rtw.ml.cmu.edu
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 3: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/3.jpg)
NELL: Never-Ending Language Learner
Inputs: l initial ontology l handful of examples of each predicate in ontology l the web l occasional interaction with human trainers
The task:
l run 24x7, forever • each day: 1. extract more facts from the web to populate the initial ontology 2. learn to read (perform #1) better than yesterday
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 4: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/4.jpg)
NELL: Never-Ending Language Learner
Goal: • run 24x7, forever • each day:
1. extract more facts from the web to populate given ontology 2. learn to read better than yesterday
Today... Running 24 x 7, since January, 2010 Input: • ontology defining ~800 categories and relations • 10-20 seed examples of each • 1 billion web pages (ClueWeb – Jamie Callan) Result: • continuously growing KB with +1,300,000 extracted beliefs
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 5: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/5.jpg)
http://rtw.ml.cmu.edu
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 6: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/6.jpg)
Bayesian Sets (BS)
Given and , rank the elements of by how well they would “fit into” a set which includes
Define a score for each : From Bayes rule, the score can be re-written as:
}{x=D DDc ⊂ D
cD
)()(
)(xx
xpDp
score c=
D∈x
)()(),()(c
c
DppDpscore
xxx =
Ghahramani & Heller; NIPS 2005
![Page 7: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/7.jpg)
Bayesian Sets (BS)
Intuitively, the score compares the probability that x and Dc were generated by the same model with the same unknown parameters θ, to the probability that x and Dc came from models with different parameters θ and θ’.
Ghahramani & Heller; NIPS 2005
)()(),()(c
c
DppDpscore
xxx =
![Page 8: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/8.jpg)
Bayesian Sets (BS)
Intuitively, the score compares the probability that x and Dc were generated by the same model with the same unknown parameters θ, to the probability that x and Dc came from models with different parameters θ and θ’.
Ghahramani & Heller; NIPS 2005
)()(),()(c
c
DppDpscore
xxx =
![Page 9: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/9.jpg)
Bayesian Sets (BS)
Intuitively, the score compares the probability that x and Dc were generated by the same model with the same unknown parameters θ, to the probability that x and Dc came from models with different parameters θ and θ’.
Ghahramani & Heller; NIPS 2005
)()(),()(c
c
DppDpscore
xxx =
![Page 10: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/10.jpg)
Bayesian Sets (BS)
Intuitively, the score compares the probability that x and Dc were generated by the same model with the same unknown parameters θ, to the probability that x and Dc came from models with different parameters θ and θ’.
Ghahramani & Heller; NIPS 2005
)()(),()(c
c
DppDpscore
xxx =
![Page 11: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/11.jpg)
BS using NELL’s Ontology
Initial ontology:
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company Vegetable Sport
![Page 12: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/12.jpg)
Initial ontology:
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport
Apple Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
BS using NELL’s Ontology
![Page 13: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/13.jpg)
Given a huge web corpus, run BS once
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
BS using NELL’s Ontology
![Page 14: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/14.jpg)
Given a huge web corpus, run BS once
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco Facebook
DELL …
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle Alan Turing
Alexander Fleming …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon Baseball
Badminton …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane Beijing Cairo …
BS using NELL’s Ontology
![Page 15: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/15.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
BS using NELL’s Ontology
![Page 16: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/16.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
BS using NELL’s Ontology
![Page 17: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/17.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
BS using NELL’s Ontology
![Page 18: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/18.jpg)
Given a huge web corpus, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud
Basketball Football
Swimming Tennis Golf
Soccer Volleyball
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Iterative BS using NELL’s Ontology Zhang & Liu, 2011
![Page 19: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/19.jpg)
Given a huge web corpus, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane
Iterative BS using NELL’s Ontology Zhang & Liu, 2011
![Page 20: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/20.jpg)
Given a huge web corpus, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco Facebook
DELL …
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle Alan Turing
Alexander Fleming …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon Baseball
Badminton …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane Beijing Cairo …
Iterative BS using NELL’s Ontology Zhang & Liu, 2011
![Page 21: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/21.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Iterative BS using NELL’s Ontology
![Page 22: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/22.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Iterative BS using NELL’s Ontology
![Page 23: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/23.jpg)
NELL: Coupled semi-supervised training of many functions
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 24: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/24.jpg)
Coupled Training Type 2: Structured Outputs, Multitask, Posterior
Regularization, Multilabel Learn functions with the same input, different outputs, where we know some constraint
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 25: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/25.jpg)
Coupled Training Type 2: Structured Outputs, Multitask, Posterior
Regularization, Multilabel Learn functions with the same input, different outputs, where we know some constraint
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 26: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/26.jpg)
Coupled Training Type 2: Structured Outputs, Multitask, Posterior
Regularization, Multilabel Learn functions with the same input, different outputs, where we know some constraint
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 27: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/27.jpg)
Coupled Bayesian Sets (CBS)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 28: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/28.jpg)
Coupled Bayesian Sets (CBS)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 29: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/29.jpg)
Coupled Bayesian Sets (CBS)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 30: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/30.jpg)
Coupled Bayesian Sets (CBS)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 31: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/31.jpg)
Coupled Bayesian Sets (CBS)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 32: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/32.jpg)
Coupled Bayesian Sets (CBS)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
![Page 33: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/33.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 34: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/34.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
MutuallyExclusive(Company,Person); MutuallyExclusive(Company,Sport); MutuallyExclusive(Company,City); MutuallyExclusive(Pearson,Sport); …
![Page 35: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/35.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
MutuallyExclusive(Company,Person); MutuallyExclusive(Company,Sport); MutuallyExclusive(Company,City); MutuallyExclusive(Pearson,Sport); …
![Page 36: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/36.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
MutuallyExclusive(Company,Person); MutuallyExclusive(Company,Sport); MutuallyExclusive(Company,City); MutuallyExclusive(Pearson,Sport); …
![Page 37: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/37.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
MutuallyExclusive(Company,Person); MutuallyExclusive(Company,Sport); MutuallyExclusive(Company,City); MutuallyExclusive(Pearson,Sport); …
![Page 38: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/38.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
MutuallyExclusive(Company,Person); MutuallyExclusive(Company,Sport); MutuallyExclusive(Company,City); MutuallyExclusive(Pearson,Sport); …
![Page 39: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/39.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
MutuallyExclusive(Company,Person); MutuallyExclusive(Company,Sport); MutuallyExclusive(Company,City); MutuallyExclusive(Pearson,Sport); …
![Page 40: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/40.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
MutuallyExclusive(Company,Person); MutuallyExclusive(Company,Sport); MutuallyExclusive(Company,City); MutuallyExclusive(Pearson,Sport); …
![Page 41: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/41.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
MutuallyExclusive(Company,Person); MutuallyExclusive(Company,Sport); MutuallyExclusive(Company,City); MutuallyExclusive(Pearson,Sport); …
![Page 42: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/42.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 43: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/43.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 44: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/44.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
… AT&T
Boeing
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 45: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/45.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
… AT&T
Boeing
Basketball Football
Swimming Tennis Golf …
AT&T Boeing
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 46: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/46.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
… AT&T
Boeing
Basketball Football
Swimming Tennis Golf …
AT&T Boeing
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town …
AT&T Boeing
CBS using NELL’s Ontology
![Page 47: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/47.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 48: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/48.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 49: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/49.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 50: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/50.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
… Brazil Telecom
Texaco
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 51: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/51.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
… Brazil Telecom
Texaco
Basketball Football
Swimming Tennis Golf …
Brazil Telecom Texaco
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 52: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/52.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
… Brazil Telecom
Texaco
Basketball Football
Swimming Tennis Golf …
Brazil Telecom Texaco
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town …
Brazil Telecom Texaco
CBS using NELL’s Ontology
![Page 53: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/53.jpg)
Given a huge web corpus and mutually exclusiveness constraints, iteratively run BS
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama
Basketball Football
Swimming Tennis Golf
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town
CBS using NELL’s Ontology
![Page 54: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/54.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 55: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/55.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 56: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/56.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 57: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/57.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 58: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/58.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 59: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/59.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 60: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/60.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 61: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/61.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 62: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/62.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 63: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/63.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 64: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/64.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 65: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/65.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 66: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/66.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology
![Page 67: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/67.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco …
Great Britain Keyboard
Pencil
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane
…
![Page 68: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/68.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco …
Great Britain Keyboard
Pencil
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane
…
Not Company
![Page 69: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/69.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco …
Great Britain Keyboard
Pencil
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane
…
Not Company
Great Britain Keyboard
Pencil
![Page 70: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/70.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco …
Great Britain Keyboard
Pencil
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane
…
Not Company
Great Britain Keyboard
Pencil
![Page 71: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/71.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco …
Great Britain Keyboard
Pencil
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane
…
Not Company
Great Britain Keyboard
Pencil
MutuallyExclusive(Company,NotCompany);
![Page 72: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/72.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco …
Great Britain Keyboard
Pencil
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane
…
Not Company
Great Britain Keyboard
Pencil
MutuallyExclusive(Company,NotCompany);
![Page 73: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/73.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
Everything
Person Company City Sport Apple
Microsoft Google
IBM Yahoo AT&T
Boeing Brazil Telecom
Texaco …
Great Britain Keyboard
Pencil
Peter Flach Bill Clinton Jeremy Lin
Adele Barak Obama Dalai Lama
Freud Tom Mitchell
Aristotle …
Basketball Football
Swimming Tennis Golf
Soccer Volleyball Jogging
Marathon …
Bristol Pittsburgh
Rio de Janeiro Tokyo
Cape Town New York London
Sao Paulo Brisbane
…
Not Company
Great Britain Keyboard
Pencil
MutuallyExclusive(Company,NotCompany);
![Page 74: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/74.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
![Page 75: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/75.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
![Page 76: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/76.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
![Page 77: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/77.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
![Page 78: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/78.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
![Page 79: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/79.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
![Page 80: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/80.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What if we do not have the mutual exclusiveness constraints?
![Page 81: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/81.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What about Semantic Relations?
![Page 82: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/82.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What about Semantic Relations?
![Page 83: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/83.jpg)
ECML/PKDD2012 Bristol, UK September, 26th, 2012
CBS using NELL’s Ontology What about Semantic Relations?
![Page 84: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/84.jpg)
Conclusions
ECML/PKDD2012 Bristol, UK September, 26th, 2012
Coupled Bayesian Sets • semi-supervised learning approach to extract category
instances (e.g. country(USA), city(New York) from web pages;
• based on the original Bayesian Sets • can outperform algorithms such as the original Bayesian
Set, the Naive Bayes classifier, the Bas-all and the coupled semi-supervised logistic regression algorithm (CPL);
• can be used to automatically generate new constraints to the set expansion task even when no mutually exclusiveness relationship is previously defined
![Page 85: Coupled Bayesian Sets Algorithm for Semi-supervised ...rtw.ml.cmu.edu/papers/ecml12-verma-slides.pdfSao Paulo Brisbane Iterative BS using NELL’s Ontology Zhang & Liu, 2011 . Given](https://reader034.fdocuments.net/reader034/viewer/2022042406/5f1f9b28c28e76120b46b01f/html5/thumbnails/85.jpg)
Acknowledgements Thanks to:
ECML/PKDD2012 audience! J Also Thanks to:
• Department of Science and Technology, Government of India under Indo-Brazil cooperation programme;
• Brazilian research agencies CAPES and CNPq;
• Zoubin Ghahramani and K.A. Heller;
• All the Read The Web group;
contact: [email protected] http://rtw.ml.cmu.edu
ECML/PKDD2012 Bristol, UK September, 26th, 2012