Search

This work describes the formalization of a word structure grammar that represents the complex morphological and morphosyntactic information embedded within the word forms of an agglutinative language (Basque), giving a comprehensive linguistic description of the main morphological phenomena, such as affixation, derivation, and composition, and also taking into account the modeling of both standard and non-standard words. We have identified the relevant issues to be addressed in the representation of such a grammar.We also present the development of Morfeus+, a tool for the analysis of unrestricted texts, testing its applicability and showing that its coverage is wide and robust, allowing the efficient processing of big volumes of data.This paper describes a mature system that has required several person/years and that tries to integrate a rigorous linguistic specification together with more practical implementation matters, such as the appropriate treatment of unknown words in unrestricted texts.

Background This work deals with Natural Language Processing applied to the clinical domain. Specifically, the work deals with a Medical Entity Recognition (MER) on Electronic Health Records (EHRs). Developing a MER system entailed heavy data preprocessing and feature engineering until Deep Neural Networks (DNNs) emerged. However, the quality of the word representations in terms of embedded layers is still an important issue for the inference of the DNNs. Goal The main goal of this work is to develop a robust MER system adapting general-purpose DNNs to cope with the high lexical variability shown in EHRs. In addition, given that EHRs tend to be scarce when there are out-domain corpora available, the aim is to assess the impact of the word representations on the performance of the MER as we move to other domains. In this line, exhaustive experimentation varying information generation methods and network parameters are crucial. Methods We adapted a general purpose sequential tagger based on Bidirectional Long-Short Term Memory cells and Conditional Random Fields (CRFs) in order to make it tolerant to high lexical variability and a limited amount of corpora. To this end, we incorporated part of speech (POS) and semantic-tag embedding layers to the word representations. Results One of the strengths of this work is the exhaustive evaluation of dense word representations obtained varying not only the domain and genre but also the learning algorithms and their parameter settings. With the proposed method, we attained an error reduction of 1.71 (5.7%) compared to the state-of-the-art even that no preprocessing or feature engineering was used. Conclusions Our results indicate that dense representations built taking word order into account leverage the entity extraction system. Besides, we found that using a medical corpus (not necessarily EHRs) to infer the representations improves the performance, even if it does not correspond to the same genre.

This work describes the initial stage of designing and implementing a modular chain of Natural Language Processing tools for Basque. The main characteristic of this chain is the deep morphosyntactic analysis carried out by the first tool of the chain and the use of these morphologically rich annotations by the following linguistic processing tools of the chain. It is designed following a modular approach, showing high ease of use of its processors. Two tools have been adapted and integrated to the chain so far, and are ready to use and freely available, namely the morphosyntactic analyzer and PoS tagger, and the dependency parser. We have evaluated these tools and obtained competitive results. Furthermore, we have tested the robustness of the tools on an extensive processing of Basque documents in various research projects.

Resolving Named Entity Disambiguation task with a small knowledge base makes the task more challenging. Concretely, we present an evaluation of the state-of-the-art methods in this task for Basque NE disambiguation based on the Basque Wikipedia. We have used MFS, VSM, ESA and UKB for linking any ambiguous surface NE form occurrence in a text with its corresponding Wikipedia entry in the Basque Wikipedia version. We have analysed their performance with different corpora and as it was expected, most of them perform worse than when using big Wikipedias such as the English version, but we think these results are more realistic for less-resourced languages. We propose a new normalization factor for ESA to minimise the effect of the knowledge base size.

This article describes the different steps in the construction of EPEC (Reference Corpus for the Processing of Basque). EPEC is a corpus of standard written Basque that has been manually tagged at different levels (morphology, surface syntax, phrases) and is currently being hand tagged at deep syntax level following the Dependency Structure-based Scheme. It is aimed to be a "reference" corpus for the development and improvement of several NLP tools for Basque. This corpus has already been used for the construction of some tools such as a morphological analyser, a lemmatiser, or a shallow syntactic analyser.

Morphological information is traditionally used to develop high quality text to speech (TTS) and automatic speech recognition (ASR) systems. The use of this information improves the naturalness and intelligibility of the TTS synthesis and provides an appropriated way to select lexical units (LU) for ASR. Basque is an agglutinative language with a complex structure inside the words and the morphological information is essential both in TTS and ASR. In this work an automatic morphological segmentation tool oriented to TTS and ASR tasks is presented.

This paper describes the representation of Basque Multiword Lexical Units and the automatic processing of Multiword Expressions. After discussing and stating which kind of multiword expressions we consider to be processed at the current stage of the work, we present the representation schema of the corresponding lexical units in a general-purpose lexical database. Due to its expressive power, the schema can deal not only with fixed expressions but also with morphosyntactically flexible constructions. It also allows us to lemmatize word combinations as a unit and yet to parse the components individually if necessary. Moreover, we describe HABIL, a tool for the automatic processing of these expressions, and we give some evaluation results. This work must be placed in a general framework of written Basque processing tools, which currently ranges from the tokenization and segmentation of single words up to the syntactic tagging of general texts.

This paper presents a new methodology, based on the classical decision trees, to get a suitable set of context dependent sublexical units for Basque Continuous Speech Recognition (CSR). The original method proposed by Bahl [1] was applied as the benchmark. Then two new features were added: a data massaging to emphasise the data and a fast and efficient Growing and Pruning algorithm for DT construction. In addition, the use of the new context dependent units to build word models was addressed. The benchmark Bahl approach gave recognition rates clearly outperforming those of context independent phone-like units. Finally the new methodology improves over the benchmark DT approach.

This paper describes the components used in the design and implementation of NLP tools for Basque. These components are based on finite state technology and are devoted to the morphological analysis of Basque, an agglutinative pre-Indo-European language. We think that our design can be interesting for the treatment of other languages. The main components developed are a general and robust morphological analyser/generator and a spelling checker/corrector for Basque named Xuxen. The analyser is a basic tool for current and future work on NLP of Basque, such as the lemmatiser/tagger Euslem, an Intranet search engine or an assistant for verse-making.

In this paper we present the results of the combination of stochastic and rule-based disambiguation methods applied to Basque language1. The methods we have used in disambiguation are Constraint Grammar formalism and an HMM based tagger developed within the MULTEXT project. As Basque is an agglutinative language, a morphological analyser is needed to attach all possible readings to each word. Then, CG rules are applied using all the morphological features and this process decreases morphological ambiguity of texts. Finally, we use the MULTEXT project tools to select just one from the possible remaining tags.Using only the stochastic method the error rate is about 14%, but the accuracy may be increased by about 2% enriching the lexicon with the unknown words. When both methods are combined, the error rate of the whole process is 3.5%. Considering that the training corpus is quite small, that the HMM model is a first order one and that Constraint Grammar of Basque language is still in progress, we think that this combined method can achieve good results, and it would be appropriate for other agglutinative languages.

This paper describes the components used in the elaboration of the commercial Xuxen spelling checker/corrector for Basque. Because Basque is a highly inflected and agglutinative language, the spelling checker/corrector has been conceived as a by-product of a general purpose morphological analyser/generator. The spelling checker/corrector performs morphological decomposition in order to check misspellings and, to correct them, uses a new strategy which combines the use of an additional two-level morphological subsystem for orthographic errors, and the recognition of correct morphemes inside the world-form during the generation of proposals for typographical errors. Due to a late process of standardization of Basque, Xuxen is intended as a useful tool for standardization purposes of present day written Basque.