## Maximum Entropy Modeling with Clausal Constraints (1997)

### Cached

### Download Links

- [www.cs.kuleuven.ac.be]
- [www.cs.kuleuven.ac.be]
- DBLP

### Other Repositories/Bibliography

Venue: | In Proceedings of the 7th International Workshop on Inductive Logic Programming |

Citations: | 37 - 1 self |

### BibTeX

@INPROCEEDINGS{Dehaspe97maximumentropy,

author = {Luc Dehaspe},

title = {Maximum Entropy Modeling with Clausal Constraints},

booktitle = {In Proceedings of the 7th International Workshop on Inductive Logic Programming},

year = {1997},

pages = {109--124},

publisher = {Springer-Verlag}

}

### Years of Citing Articles

### OpenURL

### Abstract

We present the learning system Maccent which addresses the novel task of stochastic MAximum ENTropy modeling with Clausal Constraints. Maximum Entropy method is a Bayesian method based on the principle that the target stochastic model should be as uniform as possible, subject to known constraints. Maccent incorporates clausal constraints that are based on the evaluation of Prolog clauses in examples represented as Prolog programs. We build on an existing maximum-likelihood approach to maximum entropy modeling, which we upgrade along two dimensions: (1) Maccent can handle larger search spaces, due to a partial ordering defined on the space of clausal constraints, and (2) uses a richer firstorder logic format. In comparison with other inductive logic programming systems, Maccent seems to be the first that explicitly constructs a conditional probability distribution p(CjI) based on an empirical distribution ~ p(CjI) (where p(CjI) (~p(CjI)) gives the induced (observed) probability of ...

### Citations

2121 | Building a large annotated corpus of English: The Penn Treebank - Marcus, Santorini, et al. - 1993 |

1082 | A maximum entropy approach to natural language processing - Berger, Pietra, et al. - 1996 |

1058 | EÆcient induction of logic programs - Muggleton, Feng - 1990 |

551 | Inducing features of random fields - Pietra, S, et al. - 1997 |

466 | A further note on inductive generalization - Plotkin - 1971 |

429 | Generalized iterative scaling for log-linear models - Darroch, Ratchli - 1972 |

322 | A maximum entropy part-ofspeech tagger - Ratnaparkhi - 1996 |

152 | Theories for mutagenicity: a study in first-order and feature-based induction - Srinivasan, Muggleton, et al. - 1996 |

149 | PROLQG Programming for Artificial Intelligence - Bratko - 1986 |

85 | Inductive constraint logic - Raedt, Laer - 1995 |

71 | The application of inductive logic programming to finite element mesh design - Dolaak, Muggleton - 1992 |

64 | First order jk-clausal theories are PAC-learnable - Raedt, Dzeroski - 1994 |

64 | Structural regression trees
- Kramer
- 1996
(Show Context)
Citation Context ...longs the single "positive" class. The inverse "unscaling " projection allows one to calculate predictions from the induced model. Like Fors [16], which incorporates linear regress=-=ion techniques, SRT [17]-=-, which builds structural regression trees, and C0.5 [8] which builds first order clusters, Maccent can then perform first order regression from positive data only. We applied this technique to two be... |

58 | Image reconstruction from incomplete and noisy data - Gull, Daniell - 1978 |

38 | First order regression - Karalič, Bratko - 1997 |

26 | A note on approximations of discrete probability distributions - Brown - 1959 |

23 | Naive Bayesian classifier within ILP-R - Pompe, Kononenko - 1995 |

22 | Using logical decision trees for clustering
- Raedt, Blockeel
- 1997
(Show Context)
Citation Context ... " projection allows one to calculate predictions from the induced model. Like Fors [16], which incorporates linear regression techniques, SRT [17], which builds structural regression trees, and =-=C0.5 [8]-=- which builds first order clusters, Maccent can then perform first order regression from positive data only. We applied this technique to two benchmark domains well known in the inductive logic progra... |

17 | Dlab: A declarative language bias formalism - Dehaspe, Raedt - 1996 |

14 | Induction in logic - Raedt - 1996 |

13 | Notes on present status and future prospects - Jaynes - 1991 |

9 |
Raedt. Experiments with top-down induction of logical decision trees
- Blockeel, De
- 1997
(Show Context)
Citation Context ...od approach to maximum entropy modeling, which we upgrade along two dimensions: (1) Maccent can handle larger search spaces, due to a partial ordering defined on the space of clausal constraints, and =-=(2)-=- uses a richer first-order logic format. In comparison with other inductive logic programming systems, Maccent seems to be the first that explicitly constructs a conditional probability distribution p... |

9 |
Probabilistic First-Order Classification
- Pompe, Kononenko
- 1997
(Show Context)
Citation Context ...he evaluation of Prolog clauses in examples represented as Prolog programs. The integration of probabilistic methods with inductive logic programming has recently become a popular research topic (cf. =-=[5, 16, 19, 21, 22]-=-), but, to the best of our knowledge, Maccent is the first inductive logic programming algorithm that explicitly constructs a conditional probability distribution p(CjI) based on an empirical distribu... |

4 |
Bayesian Inductive Logic Programming with explicit probabilistic bias
- Cussens
- 1996
(Show Context)
Citation Context ...he evaluation of Prolog clauses in examples represented as Prolog programs. The integration of probabilistic methods with inductive logic programming has recently become a popular research topic (cf. =-=[5, 16, 19, 21, 22]-=-), but, to the best of our knowledge, Maccent is the first inductive logic programming algorithm that explicitly constructs a conditional probability distribution p(CjI) based on an empirical distribu... |

2 |
Maximum entropy toolkit, release 1.5 beta
- Ristad
- 1997
(Show Context)
Citation Context ...propriate multi-class datasets where empirical conditional probability ~ p can take any real value between 0 and 1. For these experiments we intend to make use of the Maximum Entropy Modeling Toolkit =-=[24]-=- to improve our initial prototypical implementation. Future theoretical research should clarify the relationship between Maccent and (more) standard inductive logic programming approaches to classific... |