• Documents
  • Authors
  • Tables
  • Log in
  • Sign up
  • MetaCart
  • DMCA
  • Donate

CiteSeerX logo

Advanced Search Include Citations
Advanced Search Include Citations

A quick guide to maltparser optimization (2010)

by Joakim Nivre, Johan Hall
Add To MetaCart

Tools

Sorted by:
Results 1 - 5 of 5

MaltOptimizer: A System for MaltParser Optimization

by Miguel Ballesteros
"... Freely available statistical parsers often require careful optimization to produce state-of-the-art results, which can be a non-trivial task especially for application developers who are not interested in parsing research for its own sake. We present MaltOptimizer, a freely available tool developed ..."
Abstract - Cited by 9 (1 self) - Add to MetaCart
Freely available statistical parsers often require careful optimization to produce state-of-the-art results, which can be a non-trivial task especially for application developers who are not interested in parsing research for its own sake. We present MaltOptimizer, a freely available tool developed to facilitate parser optimization using the open-source system MaltParser, a data-driven parser-generator that can be used to train dependency parsers given treebank data. MaltParser offers a wide range of parameters for optimization, including nine different parsing algorithms, two different machine learning libraries (each with a number of different learners), and an expressive specification language that can be used to define arbitrarily rich feature models. MaltOptimizer is an interactive system that first performs an analysis of the training set in order to select a suitable starting point for optimization and then guides the user through the optimization of parsing algorithm, feature model, and learning algorithm. Empirical evaluation on data from the CoNLL 2006 and 2007 shared tasks on dependency parsing shows that MaltOptimizer consistently improves over the baseline of default settings and sometimes even surpasses the result of manual optimization.
(Show Context)

Citation Context

...ve dependencies is intermediate, both groups of algorithms are explored. In order to reduce the number of tests needed, we have came up with two different decision trees based on previous experience (=-=Nivre and Hall, 2010-=-). The first one, shown in Figure 1, tests only projective algorithms in such a way that the maximum number of tests is 3, and the procedure avoids unnecessary tests such as testing the Nivre arc-stan...

MaltOptimizer: An Optimization Tool for MaltParser

by Miguel Ballesteros, Joakim Nivre
"... Data-driven systems for natural language processing have the advantage that they can easily be ported to any language or domain for which appropriate training data can be found. However, many data-driven systems require careful tuning in order to achieve optimal performance, which may require specia ..."
Abstract - Cited by 9 (2 self) - Add to MetaCart
Data-driven systems for natural language processing have the advantage that they can easily be ported to any language or domain for which appropriate training data can be found. However, many data-driven systems require careful tuning in order to achieve optimal performance, which may require specialized knowledge of the system. We present MaltOptimizer, a tool developed to facilitate optimization of parsers developed using MaltParser, a data-driven dependency parser generator. MaltOptimizer performs an analysis of the training data and guides the user through a three-phase optimization process, but it can also be used to perform completely automatic optimization. Experiments show that MaltOptimizer can improve parsing accuracy by up to 9 percent absolute (labeled attachment score) compared to default settings. During the demo session, we will run MaltOptimizer on different data sets (user-supplied if possible) and show how the user can interact with the system and track the improvement in parsing accuracy. 1
(Show Context)

Citation Context

... added to the model. Since an exhaustive search for the best possible feature model is impossible, the system relies on a greedy optimization strategy using heuristics derived from proven experience (=-=Nivre and Hall, 2010-=-). The major steps of the forward selection experiments are the following: 4 1. Tune the window of POSTAG n-grams over the parser state. 2. Tune the window of FORM features over the parser state. 3. T...

Effective Morphological Feature Selection with MaltOptimizer at the SPMRL 2013 Shared Task

by Miguel Ballesteros
"... The inclusion of morphological features provides very useful information that helps to enhance the results when parsing morphologically rich languages. MaltOptimizer is a tool, that given a data set, searches for the optimal parameters, parsing algorithm and optimal feature set achieving the best re ..."
Abstract - Add to MetaCart
The inclusion of morphological features provides very useful information that helps to enhance the results when parsing morphologically rich languages. MaltOptimizer is a tool, that given a data set, searches for the optimal parameters, parsing algorithm and optimal feature set achieving the best results that it can find for parsers trained with MaltParser. In this paper, we present an extension of MaltOptimizer that explores, one by one and in combination, the features that are geared towards morphology. From our experiments in the context of the Shared Task on Parsing Morphologically Rich Languages, we extract an in-depth study that shows which features are actually useful for transition-based parsing and we provide competitive results, in a fast and simple way.
(Show Context)

Citation Context

.... 2 MaltOptimizer MaltOptimizer is a system written in Java that implements a full optimization procedure for MaltParser based on the experience acquired from previous experiments (Hall et al., 2007; =-=Nivre and Hall, 2010-=-). MaltOptimizer attempts to find the best model that it can find, but it does not guarantee that the outcome is the best model possible because of the difficulty of exploring all the possibilities th...

Optimizing Planar and 2-Planar Parsers with MaltOptimizer Optimizando los Parsers Planar y 2-Planar con MaltOptimizer

by Miguel Ballesteros, Carlos Gómez-rodríguez, Joakim Nivre
"... Resumen: MaltOptimizer es una herramienta capaz de proporcionar una optimización para modelos generados mediante MaltParser. Los analizadores de dependencias actuales requieren una completa configuración para obtener resultados a la altura del estado del arte, y para ello es necesario un conocimient ..."
Abstract - Add to MetaCart
Resumen: MaltOptimizer es una herramienta capaz de proporcionar una optimización para modelos generados mediante MaltParser. Los analizadores de dependencias actuales requieren una completa configuración para obtener resultados a la altura del estado del arte, y para ello es necesario un conocimiento especializado. Los analizadores Planar y 2-Planar son dos algoritmos diferentes y de reciente incorporación en MaltParser. En el presente artículo presentamos cómo estos dos analizadores pueden incluirse en MaltOptimizer comparándolos con el resto de familias de algoritmos incluidas en MaltParser, y cómo se puede definir una búsqueda y selección de atributos (o “features”) usando el propio sistema para estos dos parsers. Los experimentos muestran que usando estos métodos podemos mejorar la precisión obtenida hasta un porcentaje absoluto del 8 por ciento (labeled attachment score) si lo comparamos con una configuración básica de estos 2 parsers. Palabras clave: Análisis sintáctico de dependencias, MaltOptimizer, MaltParser, Planar y 2-Planar Abstract: MaltOptimizer is a tool that is capable of finding an optimal configuration

Exploring Automatic Feature Selection for Transition-Based Dependency Parsing

by Miguel Ballesteros
"... Resumen: En este art́ıculo se investigan técnicas automáticas para encontrar un modelo óptimo de caracteŕısticas en el caso de un analizador de dependencias basado en transiciones. Mostramos un estudio comparativo entre algoritmos de búsqueda, sistemas de validación y reglas de decisión demos ..."
Abstract - Add to MetaCart
Resumen: En este art́ıculo se investigan técnicas automáticas para encontrar un modelo óptimo de caracteŕısticas en el caso de un analizador de dependencias basado en transiciones. Mostramos un estudio comparativo entre algoritmos de búsqueda, sistemas de validación y reglas de decisión demostrando al mismo tiempo que usando nuestros métodos es posible conseguir modelos complejos que proporcionan mejores resultados que los modelos que siguen configuraciones por defecto. Palabras clave: Análisis de dependencias, MaltOptimizer, MaltParser Abstract: In this paper we investigate automatic techniques for finding an opti-mal feature model in the case of transition-based dependency parsing. We show a comparative study making a distinction between search algorithms, validation and decision rules demonstrating at the same time that using our methods it is possible to come up with quite complex feature specifications which are able to provide bet-ter results than default feature models.
(Show Context)

Citation Context

...hm is the one implemented and included in the MaltOptimizer distribution (Ballesteros and Nivre, 2012), it minimizes the number of experiments according to linguistic expert knowledge and experience (=-=Nivre and Hall, 2010-=-). It also follows the steps shown at the beginning of Section 4. However, in spite of trying with all the big set of possible features for each step as it is done in the Relaxed Greedy algorithm, it ...

Powered by: Apache Solr
  • About CiteSeerX
  • Submit and Index Documents
  • Privacy Policy
  • Help
  • Data
  • Source
  • Contact Us

Developed at and hosted by The College of Information Sciences and Technology

© 2007-2019 The Pennsylvania State University