Results 1 -
2 of
2
Multilingual deterministic dependency parsing framework using modified finite Newton method support vector machines
- In Proc. of the CoNLL 2007 Shared Task. EMNLP-CoNLL
, 2007
"... In this paper, we present a three-step multilingual dependency parser based on a deterministic shift-reduce parsing algorithm. Different from last year, we separate the root-parsing strategy as sequential labeling task and try to link the neighbor word dependences via a near neighbor parsing. The ou ..."
Abstract
-
Cited by 1 (0 self)
- Add to MetaCart
In this paper, we present a three-step multilingual dependency parser based on a deterministic shift-reduce parsing algorithm. Different from last year, we separate the root-parsing strategy as sequential labeling task and try to link the neighbor word dependences via a near neighbor parsing. The outputs of the root and neighbor parsers were encoded as features for the shift-reduce parser. In addition, the learners we used for the two parsers and the shift-reduce parser are quite different (conditional random fields and the modified finite-Newton method support vector machines). We found that our method could benefit from the two-preprocessing stages. To speed up training, in this year, we employ the MFN-SVM (modified finite-Newton method support vector machines) which can be learned in linear time. The experimental results show that our method achieved the middle rank over the 23 teams. We expect that our method could be further improved via well-tuned parameter validations for different languages. 1
†Departamento de Ingeniería del Software e Inteligencia Artificial ‡Instituto de Tecnología del Conocimiento
"... Resumen: En los últimos años los sistemas basados en aprendizaje automático desarrollados para realizar análisis sintáctico de dependencias han alcanzado una gran precisión, pero ésta está normalmente por debajo del 90 % en Labelled Attachment Score (LAS). Maltparser es un ejemplo de ese tipo de sis ..."
Abstract
- Add to MetaCart
Resumen: En los últimos años los sistemas basados en aprendizaje automático desarrollados para realizar análisis sintáctico de dependencias han alcanzado una gran precisión, pero ésta está normalmente por debajo del 90 % en Labelled Attachment Score (LAS). Maltparser es un ejemplo de ese tipo de sistemas. El aprendizaje automático permite obtener analizadores para cada lengua para la que se disponga de un corpus de entrenamiento adecuado. Dado que generalmente tales sistemas no pueden ser modificados, surge la siguiente cuestión: ¿Se puede mejorar el 90 % en LAS utilizando mejores corpora de entrenamiento? En este artículo describimos trabajos prospectivos sobre la cuestión, estudiando estrategias en las que se consideran tanto el tamaño del corpus como las longitudes de sus frases con el fin de obtener una mejor precisión en el análisis. Palabras clave: Análisis sintáctico de dependencias, Maltparser, español, precisión Abstract: In the last years, dependency parsing has been accomplished by machine learning–based systems showing great accuracy but usually under 90 % for Labelled Attachment Score (LAS). Maltparser is one of such systems. Machine learning allows to obtain parsers for every language having an adequate training corpus. Since generally such systems can not be modified the following question arises: Can we beat this 90 % LAS by using better training corpora? In the present paper we show some prospective works on it. We studied some strategies considering training corpus ’ size and its sentences ’ length in order to obtain better parsing accuracy.

