Search

The objective of this work is to set a corpus-driven methodology to quantify automatically diachronic language distance between chronological periods of several languages. We apply a perplexity-based measure to written text representing different historical periods of three languages: European English, European Portuguese, and European Spanish. For this purpose, we have built historical corpora for each period, which have been compiled from different open corpus sources containing texts as close as possible to its original spelling. The results of our experiments show that a diachronic language distance based on perplexity detects the linguistic evolution that had already been explained by the historians of the three languages. It is remarkable to underline that it is an unsupervised multilingual method which only needs a raw corpora organized by periods.

Neste artigo propomo-nos a definir e desenvolver uma estratégia automática para procurar especificidades lexicais dentro de conjuntos de textos utilizando unidades lexicais simples e expressões com várias palavras, ou termos multipalavra (MWE, a sua sigla em inglês). Propomos uma metodologia para o cálculo da divergência de distribuições de lemas e de MWE que permitirá encontrar, automaticamente, diferenças e semelhanças entre textos não anotados. Esta metodologia poderá ser utilizada para posteriormente identificar grupos de textos sobre os quais se procederá a análises quantitativas e qualitativas semiautomáticas e/ou com intervenção humana. Num primeiro teste, utilizamos dois textos de especialidade (da área da pediatria) e um texto literário, presumindo que os textos de especialidade deveriam apresentar maiores divergências relativamente ao texto literário do que entre eles próprios. Como os testes feitos mostraram a tendência esperada, decidimos aplicar a mesma metodologia a um segundo grupo de textos (três conjuntos de entrevistas a visitantes da cidade de Santiago de Compostela)., Este trabalho é apoiado pelo projeto Narrativas, usos e consumos de visitantes como aliados ou amena¸cas para o bem-estar da comunidade local: o caso de Santiago de Compostela. Ref: FFI2017-88196-R, parcialmente subsidiado pelo Ministerio de Industria, Econom´ıa y Competitividad espanhol no quadro do Programa Estatal de I+D+i Orientada a los Retos de la Sociedad (2018-2021).

In this paper, we describe a grammatical formalism, called DepPattern, to write dependency grammars using patterns of Part of Speech (PoS) tags augmented with lexical and morphological information. The formalism inherits ideas from Sinclair’s work and Pattern Grammar. To properly analyze semi-fixed idiomatic expressions, DepPattern distinguishes between open-choice and idiomatic rules. A grammar is defined as a set of lexical-syntactic rules at different levels of abstraction. In addition, a compiler was implemented so as to generate deterministic and robust parsers from DepPattern grammars. These parsers identify dependencies which can be used to improve corpus-based applications such as information extraction. At the end of this article, we describe an experiment which evaluates the efficiency of a dependency parser generated from a simple DepPattern grammar. In particular, we evaluated the precision of a semantic extraction method making use of a DepPattern-based parser.

We describe the role of morphosyntactic categories and syntactic dependencies in the process of semantically interpreting composite expressions. Special attention will be paid to the combinatorial properties conveyed by morphosyntactic categories such as nominals and verbs, as well as by syntactic dependencies like subject, direct object, or nominal modification. The semantic characterisation of these grammatical structures is based on cognitive abilities and abstract conceptualisations. This will provide us with theoretical arguments to review and extend some basic assumptions of Cognitive Grammar and related work.

This paper outlines a strategy to build new bilingual dictionaries from existing resources. The method is based on two main tasks: first, a new set of bilingual correspondences is generated from two available bilingual dictionaries. Second, the generated correspondences are validated by making use of a bilingual lexicon automatically extracted from non-parallel, and comparable corpora. The quality of the entries of the derived dictionary is very high, similar to that of hand-crafted dictionaries. We report a case study where a new, non noisy, English-Galician dictionary with about 12,000 correct bilingual correspondences was automatically generated.

So far, research on extraction of translation equivalents from comparable, non-parallel corpora has not been very popular. The main reason was the poor results when compared to those obtained from aligned parallel corpora. The method proposed in this paper, relying on seed patterns generated from external bilingual dictionaries, allows us to achieve similar results to those from parallel corpus. In this way, the huge amount of comparable corpora available via Web can be viewed as a never-ending source of lexicographic information. In this paper, we describe the experiments performed on a comparable, Spanish-Galician corpus.

Descripción del proyecto Gari-Coter para la elaboración de los recursos lingüísticos en gallego necesarios para un re-elaborador de consultas multilingüe. Description of the Gari-Coter project for the development of the necessary linguistic resources in Galician for a multilingual query re-elaborator.

Few approaches to extract word translations from non-parallel texts have been proposed so far. Researchers have not been encouraged to work on this topic because extracting information from non-parallel corpora is a difficult task producing poor results. Whereas for parallel texts, word translation extraction can reach about 99%, the accuracy for non-parallel texts has been around 72% up to now. The current approach, which relies on the previous extraction of bilingual pairs of lexico-syntactic templates from parallel corpora, makes a significant improvement to about 89% of words translations identified correctly.