Results 1  10
of
98
Markov Logic Networks
 Machine Learning
, 2006
"... Abstract. We propose a simple approach to combining firstorder logic and probabilistic graphical models in a single representation. A Markov logic network (MLN) is a firstorder knowledge base with a weight attached to each formula (or clause). Together with a set of constants representing objects ..."
Abstract

Cited by 569 (34 self)
 Add to MetaCart
Abstract. We propose a simple approach to combining firstorder logic and probabilistic graphical models in a single representation. A Markov logic network (MLN) is a firstorder knowledge base with a weight attached to each formula (or clause). Together with a set of constants representing objects in the domain, it specifies a ground Markov network containing one feature for each possible grounding of a firstorder formula in the KB, with the corresponding weight. Inference in MLNs is performed by MCMC over the minimal subset of the ground network required for answering the query. Weights are efficiently learned from relational databases by iteratively optimizing a pseudolikelihood measure. Optionally, additional clauses are learned using inductive logic programming techniques. Experiments with a realworld database and knowledge base in a university domain illustrate the promise of this approach.
Enhanced Hypertext Categorization Using Hyperlinks
, 1998
"... A major challenge in indexing unstructured hypertext databases is to automatically extract metadata that enables structured search using topic taxonomies, circumvents keyword ambiguity, and improves the quality of search and profilebased routing and filtering. Therefore, an accurate classifier is ..."
Abstract

Cited by 383 (8 self)
 Add to MetaCart
A major challenge in indexing unstructured hypertext databases is to automatically extract metadata that enables structured search using topic taxonomies, circumvents keyword ambiguity, and improves the quality of search and profilebased routing and filtering. Therefore, an accurate classifier is an essential component of a hypertext database. Hyperlinks pose new problems not addressed in the extensive text classification literature. Links clearly contain highquality semantic clues that are lost upon a purely termbased classifier, but exploiting link information is nontrivial because it is noisy. Naive use of terms in the link neighborhood of a document can even degrade accuracy. Our contribution is to propose robust statistical models and a relaxation labeling technique for better classification by exploiting link information in a small neighborhood around documents. Our technique also adapts gracefully to the fraction of neighboring documents having known topics. We experimented ...
Probabilistic models of relational structure
 In Proc. ICML
, 2001
"... Most realworld data is stored in relational form. In contrast, most statistical learning methods work with “flat ” data representations, forcing us to convert our data into a form that loses much of the relational structure. The recently introduced framework of probabilistic relational models (PRMs ..."
Abstract

Cited by 104 (11 self)
 Add to MetaCart
Most realworld data is stored in relational form. In contrast, most statistical learning methods work with “flat ” data representations, forcing us to convert our data into a form that loses much of the relational structure. The recently introduced framework of probabilistic relational models (PRMs) allows us to represent probabilistic models over multiple entities that utilize the relations between them. In this paper, we propose the use of probabilistic models not only for the attributes in a relational model, but for the relational structure itself. We propose two mechanisms for modeling structural uncertainty: reference uncertainty and existence uncertainty. We describe the appropriate conditions for using each model and present learning algorithms for each. We present experimental results showing that the learned models can be used to predict relational structure and, moreover, the observed relational structure can be used to provide better predictions for the attributes in the model. 1.
Relational Reinforcement Learning
, 2001
"... Relational reinforcement learning is presented, a learning technique that combines reinforcement learning with relational learning or inductive logic programming. Due to the use of a more expressive representation language to represent states, actions and Qfunctions, relational reinforcement learni ..."
Abstract

Cited by 102 (6 self)
 Add to MetaCart
Relational reinforcement learning is presented, a learning technique that combines reinforcement learning with relational learning or inductive logic programming. Due to the use of a more expressive representation language to represent states, actions and Qfunctions, relational reinforcement learning can be potentially applied to a new range of learning tasks. One such task that we investigate is planning in the blocks world, where it is assumed that the effects of the actions are unknown to the agent and the agent has to learn a policy. Within this simple domain we show that relational reinforcement learning solves some existing problems with reinforcement from specific goals pursued and to exploit the results of previous learning phases when addressing new (more complex) situations.
Learning Probabilistic Models of Link Structure
 Journal of Machine Learning Research
, 2002
"... Most realworld data is heterogeneous and richly interconnected. Examples include the Web, hypertext, bibliometric data and social networks. In contrast, most statistical learning methods work with "flat" data representations, forcing us to convert our data into a form that loses much of the link ..."
Abstract

Cited by 102 (12 self)
 Add to MetaCart
Most realworld data is heterogeneous and richly interconnected. Examples include the Web, hypertext, bibliometric data and social networks. In contrast, most statistical learning methods work with "flat" data representations, forcing us to convert our data into a form that loses much of the link structure. The recently introduced framework of probabilistic relational models (PRMs) embraces the objectrelational nature of structured data by capturing probabilistic interactions between attributes of related entities. In this paper, we extend this framework by modeling interactions between the attributes and the link structure itself. An advantage of our approach is a unified generarive model for both content and relational structure. We propose two mechanisms for representing a probabilistic distribution over link structures: reference uncertainty and existence uncertainty. We describe the appropriate conditions for using each model and present learning algorithms for each. We present experimental results showing that the learned models can be used to predict link structure and, moreover, the observed link structure can be used to provide better predictions for the attributes in the model.
Learning the structure of Markov logic networks
 In Proceedings of the 22nd International Conference on Machine Learning
, 2005
"... Markov logic networks (MLNs) combine logic and probability by attaching weights to firstorder clauses, and viewing these as templates for features of Markov networks. In this paper we develop an algorithm for learning the structure of MLNs from relational databases, combining ideas from inductive l ..."
Abstract

Cited by 88 (17 self)
 Add to MetaCart
Markov logic networks (MLNs) combine logic and probability by attaching weights to firstorder clauses, and viewing these as templates for features of Markov networks. In this paper we develop an algorithm for learning the structure of MLNs from relational databases, combining ideas from inductive logic programming (ILP) and feature induction in Markov networks. The algorithm performs a beam or shortestfirst search of the space of clauses, guided by a weighted pseudolikelihood measure. This requires computing the optimal weights for each candidate structure, but we show how this can be done efficiently. The algorithm can be used to learn an MLN from scratch, or to refine an existing knowledge base. We have applied it in two realworld domains, and found that it outperforms using offtheshelf ILP systems to learn the MLN structure, as well as pure ILP, purely probabilistic and purely knowledgebased approaches. 1.
Induction of logic programs: FOIL and related systems
 NEW GENERATION COMPUT
, 1995
"... FOIL is a firstorder learning system that uses information in a collection of relations to construct theories expressed in a dialect of Prolog. This paper provides an overview of the principal ideas and methods used in the current version of the system, including two recent additions. We present ex ..."
Abstract

Cited by 61 (1 self)
 Add to MetaCart
FOIL is a firstorder learning system that uses information in a collection of relations to construct theories expressed in a dialect of Prolog. This paper provides an overview of the principal ideas and methods used in the current version of the system, including two recent additions. We present examples of tasks tackled by foil and of systems that adapt and extend its approach.
Understanding the crucial role of attribute interaction in data mining
 Artif. Intel. Rev
, 2001
"... This is a review paper, whose goal is to significantly improve our understanding of the crucial role of attribute interaction in data mining. The main contributions of this paper are as follows. Firstly, we show that the concept of attribute interaction has a crucial role across different kinds of p ..."
Abstract

Cited by 48 (14 self)
 Add to MetaCart
This is a review paper, whose goal is to significantly improve our understanding of the crucial role of attribute interaction in data mining. The main contributions of this paper are as follows. Firstly, we show that the concept of attribute interaction has a crucial role across different kinds of problem in data mining, such as attribute construction, coping with small disjuncts, induction of firstorder logic rules, detection of Simpson’s paradox, and finding several types of interesting rules. Hence, a better understanding of attribute interaction can lead to a better understanding of the relationship between these kinds of problems, which are usually studied separately from each other. Secondly, we draw attention to the fact that most rule induction algorithms are based on a greedy search which does not cope well with the problem of attribute interaction, and point out some alternative kinds of rule discovery methods which tend to cope better with this problem. Thirdly, we discussed several algorithms and methods for discovering interesting knowledge that, implicitly or explicitly, are based on the concept of attribute interaction.
Bottomup learning of Markov logic network structure
 In Proceedings of the TwentyFourth International Conference on Machine Learning
, 2007
"... Markov logic networks (MLNs) are a statistical relational model that consists of weighted firstorder clauses and generalizes firstorder logic and Markov networks. The current stateoftheart algorithm for learning MLN structure follows a topdown paradigm where many potential candidate structures a ..."
Abstract

Cited by 47 (6 self)
 Add to MetaCart
Markov logic networks (MLNs) are a statistical relational model that consists of weighted firstorder clauses and generalizes firstorder logic and Markov networks. The current stateoftheart algorithm for learning MLN structure follows a topdown paradigm where many potential candidate structures are systematically generated without considering the data and then evaluated using a statistical measure of their fit to the data. Even though this existing algorithm outperforms an impressive array of benchmarks, its greedy search is susceptible to local maxima or plateaus. We present a novel algorithm for learning MLN structure that follows a more bottomup approach to address this problem. Our algorithm uses a “propositional ” Markov network learning method to construct “template” networks that guide the construction of candidate clauses. Our algorithm significantly improves accuracy and learning time over the existing topdown approach in three realworld domains. 1.
Supervised versus multiple instance learning: An empirical comparison
 Proceedings of 22nd International Conference on Machine Learning (ICML2005
, 2005
"... We empirically study the relationship between supervised and multiple instance (MI) learning. Algorithms to learn various concepts have been adapted to the MI representation. However, it is also known that concepts that are PAClearnable with onesided noise can be learned from MI data. A relevant q ..."
Abstract

Cited by 45 (2 self)
 Add to MetaCart
We empirically study the relationship between supervised and multiple instance (MI) learning. Algorithms to learn various concepts have been adapted to the MI representation. However, it is also known that concepts that are PAClearnable with onesided noise can be learned from MI data. A relevant question then is: how well do supervised learners do on MI data? We attempt to answer this question by looking at a cross section of MI data sets from various domains coupled with a number of learning algorithms including Diverse Density, Logistic Regression, nonlinear Support Vector Machines and FOIL. We consider a supervised and MI version of each learner. Several interesting conclusions emerge from our work: (1) no MI algorithm is superior across all tested domains, (2) some MI algorithms are consistently superior to their supervised counterparts, (3) using high falsepositive costs can improve a supervised learner’s performance in MI domains, and (4) in several domains, a supervised algorithm is superior to any MI algorithm we tested. 1.