Learning (k,l)-context-sensitive probabilistic grammars with nonparametric Bayesian approach

An Application of the Variational Bayesian Approach to Probabilistic Context-Free Grammars

We present an efficient learning algorithm for probabilistic context-free grammars based on the variational Bayesian approach. Although the maximum likelihood method has traditionally been used for learning probabilistic language models, Bayesian learning is, in principle, less likely to cause overfitting problems than the maximum likelihood method. We show that the computational complexity of ...

Conditional Probabilistic Context-Free Grammars

Generalization in Context Sensitive Grammars

An alternative to the recognition systems used so far can be systems based on context sensitive grammars. A precondition for success will be an implementation with great capacity for generalization. We propose three methods of generalization based on the assumption that similar stimuli invoke similar or identical reactions.

Weakly Growing Context-Sensitive Grammars

This paper introduces weakly growing context-sensitive grammars. Abstract-1 Such grammars generalize the class of growing context-sensitive grammars (studied by several authors), in that these grammars have rules that “grow” according to a position valuation. If a position valuation coincides with the initial part of an expoAbstract-2 nential function, it is called a steady position valuation. ...

Attributed Context-Sensitive Graph Grammars

The paper introduces a concept of attributed context-sensitive graph grammars. The graph grammars are a graphical generalization of the textual grammars and can thus be used to specify the syntax of graphical programming or modeling languages. The attributed graph grammars extend the basic graph grammars with definitions of attributes and the associated attribute evaluation rules. By analogy to...