539 research outputs found
Sort by
Scientific Report on Rich Tree-Based SMT
No full textThe final technical report on rich tree-based translation for the EuroMatrixPlus project
CUNI: Feature Selection and Error Analysis of a Transition-Based Parser
No full textThe paper describes our system used for dependency parsing of Hindi data during the shared task at Coling 2012
Using an SVM Ensemble System for Improved Tamil Dependency Parsing
No full textDependency parsing has been shown to improve NLP systems in certain languages and in many cases helps achieve state of the art results in NLP applications, in particular applications for free word order languages. Morphologically rich languages are often short on training data or require much higher amounts of training data due to the increased size of their lexicon. This paper examines a new approach for addressing morphologically rich languages with little training data to start.
Using Tamil as our test language, we create 9 dependency parse models with a limited amount of training data. Using these models we train an SVM classifier using only the model agreements as features. We use this SVM classifier on an edge by edge decision to form an ensemble parse tree. Using only model agreements as features allows this method to remain language independent and applicable to a wide range of morphologically rich languages.
We show a statistically significant 5.44% improvement over the average dependency model and a statistically significant 0.52% improvement over the best individual system
Yet Another Language Identifier
No full textLanguage identification of written text has been studied for several decades. Despite this fact, most of the research is focused on a few most spoken languages, whereas the minor ones are ignored. The identification of a larger number of languages brings new difficulties that do not occur for a few languages. These difficulties are causing decreased accuracy. The objective of this paper is to investigate the sources of such degradation. In order to isolate the impact of individual factors, 5 different algorithms and 3 different number of languages are used. The Support Vector Machine algorithm achieved an accuracy of 98% for 90 languages and the YALI algorithm based on a scoring function had an accuracy of 95.4%. The YALI algorithm has slightly lower accuracy but classifies around 17 times faster and its training is more than 4000 times faster.
Three different data sets with various number of languages and sample sizes were prepared to overcome the lack of standardized data sets. These data sets are now publicly available
The Czech Language in the Digital Age
No full textThis white paper presents the state of language technology support for the Czech language. It is a part of a series that analyses the available language resources and technologies for 30 European languages
Towards a Predicate-Argument Evaluation for MT
No full textHMEANT (Lo and Wu, 2011a) is a manual MT evaluation technique that focuses on
predicate-argument structure of the sentence.
We relate HMEANT to an established lin-
guistic theory, highlighting the possibilities of
reusing existing knowledge and resources for
interpreting and automating HMEANT. We
apply HMEANT to a new language, Czech
in particular, by evaluating a set of English-
to-Czech MT systems. HMEANT proves to
correlate with manual rankings at the sentence
level better than a range of automatic met-
rics. However, the main contribution of this
paper is the identification of several issues
of HMEANT annotation and our proposal on
how to resolve them
DEPFIX: A System for Automatic Correction of Czech MT Outputs
Get PDFWe present an improved version of DEPFIX, a system for automatic rule-based post-processing of English-to-Czech MT outputs designed to increase their fluency. We enhanced the rule set used by the original DEPFIX system and measured
the performance of the individual rules.
We also modified the dependency parser of
McDonald et al. (2005) in two ways to adjust
it for the parsing of MT outputs. We show that
our system is able to improve the quality of the
state-of-the-art MT systems
Wild Experimenting in Machine Translation
No full textA tutorial on conducting a very broad range of experiments (for MT in particular)
Treex::Core
No full textTreex::Core is a set of Perl modules aimed at a general processing of syntactic tree structures. The data structures are designed to be as language independent as possible. Treex::Core is currently used especially for developing syntax-based Machine Translation systems
HamleDT - HArmonized Multi-LanguagE Dependency Treebank
No full textHamleDT is a set of software tools for conversion (hamonizing) of treebanks, which were created for different languages and based on various linguistic formalisms, into the same annotation framework. The main aim is to facilitate development of multilingual technology. HamleDT covers as many as 28 languages in its current version