Grammatical inference: algorithms and applications

En savoir plus sur le livre

The book features a comprehensive exploration of grammatical inference and its applications across various domains. It includes invited papers on parsing without grammar rules and the classification of biological sequences using kernel methods. Regular papers delve into topics such as the identification of systematic-noisy languages, ten open problems in grammatical inference, and polynomial-time identification of grammar extensions from positive data. The work also discusses PAC-learning of unambiguous NTS languages and incremental learning of context-free grammars. Further contributions address variational Bayesian grammar induction, stochastic analysis of enhanced language models, and the use of pseudo-stochastic rational languages in probabilistic inference. Topics on learning analysis, inferring grammars for mildly context-sensitive languages, and planar languages are also covered. The text highlights practical applications, such as protein motif prediction and grammatical inference in the biomedical domain, alongside challenges like inferring programming language dialects and participation in the Tenjinno Machine Translation Competition. Additionally, it presents large-scale inference of deterministic transductions, discriminative models of stochastic edit distance, and learning tree transducers. Methods for learning finite-state machines and employing MDL for grammar induction are discussed, along with merging state