Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 91
Encontrada(s) 1 página(s)

Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Towards a User-Friendly Platform for Building Language Resources based on Web Services

  • Poch, Marc
  • Toral, Antonio
  • Hamon, Olivier
  • Quochi, Valeria
  • Bel Rafecas, Núria
This paper presents the platform developed in the PANACEA project, a distributed factory that automates the stages involved in the acquisition, production, updating and maintenance of Language Resources required by Machine Translation and other Language Technologies. We adopt a set of tools that have been successfully used in the Bioinformatics field, they are adapted to the needs of our field and used to deploy web services, which can be combined to build more complex processing chains (workflows). This paper describes the platform and its different components (web services, registry, workflows, social network and interoperability). We demonstrate the scalability of the platform by carrying out a set of massive data experiments. Finally, a validation of the platform across a set of required criteria proves its usability for different types of users (non-technical users and providers).
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Efficiency-based evaluation of aligners for industrial applications

  • Toral, Antonio
  • Poch, Marc
  • Pecina, Pavel
  • Thurmair, Gregor
This paper presents a novel efficiencybased evaluation of sentence and word aligners. This assessment is critical in order to make a reliable use in industrial scenarios. The evaluation shows that the resources/nrequired by aligners differ rather broadly. Subsequently, we establish limitation mechanisms on a set of aligners deployed as web services. These results, paired with the quality expected from the aligners, allow providers to choose the most appropriate aligner according to the task at hand.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Automatic Lexical Semantic Classification of Nouns

  • Bel Rafecas, Núria
  • Romeo, Lauren
  • Padró, Muntsa
The work we present here addresses cue-based noun classification in English and Spanish. Its main objective is to automatically acquire lexical semantic information by classifying nouns into previously known noun lexical classes. This is achieved by using particular aspects of linguistic contexts as cues that identify a specific lexical class. Here we concentrate on the task of identifying such cues and the theoretical background that allows for an assessment of the complexity of the task. The results show that, despite of the a-priori complexity of the task, cue-based classification is a useful tool in the automatic acquisition of lexical semantic classes.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Using qualia information to identify lexical semantic classes in an unsupervised clustering task

  • Romeo, Lauren
  • Mendes, Sara
  • Bel Rafecas, Núria
Acquiring lexical information is a complex problem, typically approached by relying on a number of contexts to contribute information for classification. One of the first issues to address in this domain is the determination of such contexts. The work presented here proposes the use of automatically obtained FORMAL role descriptors as features used to draw nouns from the same lexical semantic class together in an unsupervised clustering task. We have dealt with three lexical semantic classes (HUMAN, LOCATION and EVENT) in English. The results obtained show that it is possible to discriminate between elements from different lexical semantic classes using only FORMAL role information, hence validating our initial hypothesis. Also, iterating our method accurately accounts for fine-grained distinctions within lexical classes, namely distinctions involving ambiguous expressions. Moreover, a filtering and bootstrapping strategy employed in extracting FORMAL role descriptors proved to minimize effects of sparse data and noise in our task.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Web Services for Bayesian Learning

  • Padró, Muntsa
  • Bel Rafecas, Núria
In this demonstration we present our web services to perform Bayesian learning for classification tasks.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Finding dependency parsing limits over a large Spanish Corpus

  • Padró, Muntsa
  • Ballesteros, Miguel
  • Martínez, Héctor
  • Bohnet, Bernd
This paper studies the performance of different parsers over a large Spanish treebank. The aim of this work is to assess the limitations of state-of-the-art parsers. We want to select the most appropriate parser for subcategorization Frame acquisition, and we focus our analysis on two aspects: the accuracy drop when parsing out-of-domain data, and the performance over specific labels relevant to our task., HM has been partially funded by the European Commission’s 7th Framework Program under grant agreement 238405 (CLARA). MP has been funded by the European Project PANACEA (FP7-ICT-2010- 248064).
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Automatic Extraction of Polar Adjectives for the Creation of Polarity Lexicons

  • Vázquez, Silvia
  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Gonzalo, Julio
Automatic creation of polarity lexicons is a crucial issue to be solved in order to reduce time and/nefforts in the first steps of Sentiment Analysis. In this paper we present a methodology based on/nlinguistic cues that allows us to automatically discover, extract and label subjective adjectives/nthat should be collected in a domain-based polarity lexicon. For this purpose, we designed a/nbootstrapping algorithm that, from a small set of seed polar adjectives, is capable to iteratively/nidentify, extract and annotate positive and negative adjectives. Additionally, the method/nautomatically creates lists of highly subjective elements that change their prior polarity even/nwithin the same domain. The algorithm proposed reached a precision of 97.5% for positive/nadjectives and 71.4% for negative ones in the semantic orientation identification task.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Towards the Fully Automatic Merging of Lexical Resources: A Step Forward

  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Necşulescu, Silvia
This article reports on the results of the research done towards the fully automatically merging of lexical resources. Our main goal is to show the generality of the proposed approach, which have been previously applied to merge Spanish Subcategorization Frames lexica. In this work we extend and apply the same technique to perform the merging of morphosyntactic lexica encoded in LMF. The experiments showed that the technique is general enough to obtain good results in these two different tasks which is an important step towards performing the merging of lexical resources fully automatically.
Repositorio Digital de la UPF
Publicación Artículo científico (JournalArticle)

Tradução online

  • Poch, Marc
  • Bel Rafecas, Núria
  • Toral, Antonio
  • Hamon, Olivier
  • Quochi, Valeria
Um desafio estratégico no mundo globalizado atual é superar a barreira da língua através da tecnologia. Na Europa, um projeto está avaliando e desenvolvendo uma plataforma que poderá traduzir a enorme quantidade de dados (orais e escritos) produzidos pelas pessoas, num processo chamado "tradução online". Conheça a seguir os detalhes do estudo.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Language Resources Factory: case study on the acquisition of Translation Memories

  • Poch, Marc
  • Toral, Antonio
  • Bel Rafecas, Núria
This paper demonstrates a novel distributed architecture to facilitate the acquisition of Language Resources. We build a factory that automates the stages involved in the acquisition, production, updating and maintenance of these resources. The factory is designed as a platform where functionalities are deployed as web services, which can be combined in complex acquisition chains using workflows. We show a case study, which acquires a Translation Memory for a given pair of languages and a domain using web services for crawling, sentence alignment and conversion to TMX.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

A Classification of Adjectives for Polarity Lexicons Enhancement

  • Vázquez, Silvia
  • Bel Rafecas, Núria
Subjective language detection is one of the most important challenges in Sentiment Analysis. Because of the weight and frequency in opinionated texts, adjectives are considered a key piece in the opinion extraction process. These subjective units are more and more frequently collected in polarity lexicons in which they appear annotated with their prior polarity. However, at the moment, any polarity lexicon takes into account prior polarity variations across domains. This paper proves that a majority of adjectives change their prior polarity value depending on the domain. We propose a distinction between domain dependent and romain independent adjectives. Moreover, our analysis led us to propose a further classification related to subjectivity degree: constant, mixed and highly subjective adjectives. Following this classification, polarity values will be a better support for Sentiment Analysis.
Repositorio Digital de la UPF
Publicación Artículo científico (JournalArticle)

Platform for Automatic, Normalized Annotation and Cost-Effective Acquisition of Language Resources for Human Language Technologies: PANACEA

  • Bel Rafecas, Núria
El objetivo de PANACEA es engranar diferentes herramientas avanzadas para construir una fábrica de Recursos Lingüísticos (RL), una línea de producción que automatice los pasos implicados en la adquisición, producción, actualización y mantenimiento de los RL que la Traducción Automática y otras tecnologías lingüísticas, necesitan.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Interoperability and Technology for a Language Resources Factory

  • Poch, Marc
  • Bel Rafecas, Núria
This document describes some of the technological aspects of a project devoted to the creation of a factory for language resources. The project’s objectives are explained, as well as the idea to create a distributed infrastructure of web services. This document focuses on two main topics of the factory: (1) the technological approaches chosen to develop the factory, i.e. software, protocols, servers, etc. (2) and Interoperability as the main challenge is to permit different NLP tools work together in the factory. This document explains why XCES and GrAF are chosen as the main formats used for the linguistic data exchange., This research has been funded by the PANACEA project (EU-7FP-ITC-248064).
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Mining and exploiting domain-specific corpora in the PANACEA platform

  • Bel Rafecas, Núria
  • Papavassiliou, Vassilis
  • Prokopidis, Prokopis
  • Toral, Antonio
  • Arranz, Victoria
The objective of the PANACEA ICT-2007.2.2 EU project is to build a platform that automates the stages involved in the acquisition,production, updating and maintenance of the large language resources required by, among others, MT systems. The development of a Corpus Acquisition Component (CAC) for extracting monolingual and bilingual data from the web is one of the most innovative building blocks of PANACEA. The CAC, which is the first stage in the PANACEA pipeline for building Language Resources, adopts an efficient and distributed methodology to crawl for web documents with rich textual content in specific languages and predefined domains. The CAC includes modules that can acquire parallel data from sites with in-domain content available in more than one language. In order to extrinsically evaluate the CAC methodology, we have conducted several experiments that used crawled parallel corpora for the identification and extraction of parallel sentences using sentence alignment. The corpora were then successfully used for domain adaptation of Machine Translation Systems.
Repositorio Digital de la UPF
Publicación Artículo científico (JournalArticle)

Verb SCF extraction for Spanish with dependency parsing = Extracción de patrones de subcategorización de verbos en castellano con análisis de dependencias

  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Garí Soler, Aina
En este artículo presentamos los resultados de nuestros experimentos en producción automática de léxicos con información de patrones de subcategorización verbal para castellano. La investigación se llevó a cabo en el marco del proyecto PANACEA de adquisición automática de información léxica que redujera al máximo la intervención humana. En nuestros experimentos, se utilizó una cadena de diferentes herramientas que incluía ‘crawling’ de textos de un dominio particular, normalización y limpieza de los textos, segmentación, identificación de unidades, etiquetado categorial y análisis de dependencias antes de, finalmente, la extracción de los patrones de subcategorización. Los resultados obtenidos muestran una gran dependencia de la calidad de los analizadores de dependencias aunque, no obstante, están en línea con los resultados obtenidos en experimentos similares para otras lenguas., In this paper we present the results of our experiments in automatic production of verb subcategorization frame lexica for Spanish. The work was carried out in the framework of the PANACEA project aiming at the automatic acquisition of lexical information reducing at maximum human intervention. In our experiments, a chain of different tools was used: domain focused web crawling, automatic cleaning, segmentation and tokenization, PoS tagging, dependency parsing and finally SCFs extraction. The obtained results show a high dependency on the quality of the results of the intervening components, in particular of the dependency parsing, which is the focus of this paper. Nevertheless, the results achieved are in line with the state-of-the-art for other languages in similar experiments., This work was funded by the European Project PANACEA (FP7-ICT-2010-248064).
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Automatic Detection of Non-deverbal Event Nouns for Quick Lexicon Production

  • Bel Rafecas, Núria
  • Coll, Maria
  • Resnik, Gabriela
In this work we present the results of experimental work on the development of lexical class-based lexica by automatic means. Our purpose is to assess the use of linguistic lexical-class based information as a feature selection methodology for the use of classifiers in quick lexical development. The results show that the approach can help reduce the human effort required in the development of language resources significantly.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

PANACEA (Platform for Automatic, Normalised Annotation and Cost-Effective Acquisition of Language Resources for Human Language Technologies)

  • Bel Rafecas, Núria
  • Poch, Marc
  • Toral, Antonio
The objective of PANACEA is to build a factory of LRs that automates the stages involved in the acquisition, production, updating and maintenance of LRs required by MT systems and by other applications based on language technologies, and simplifies eventual issues regarding intellectual property rights. This automation will cut down the cost, time and human effort significantly. These reductions of costs and time are the only way to guarantee the continuous supply of LRs that MT and other language technologies will be demanding in the multilingual Europe.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Towards the automatic merging of language resources

  • Necşulescu, Silvia
  • Bel Rafecas, Núria
  • Padró, Muntsa
  • Marimon, Montserrat
  • Revilla, Eva
Language Resources are a critical component for Natural Language Processing applications. Throughout the years many resources were manually created for the same task, but with different granularity and coverage information. To create richer resources for a broad range of potential reuses, nformation from all resources has to be joined into one. The hight cost of comparing and merging different resources by hand has been a bottleneck for merging existing resources. With the objective of reducing human intervention, we present a new method for automating merging resources. We have addressed the merging of two verbs subcategorization frame (SCF) lexica for Spanish. The results achieved, a new lexicon with enriched information and conflicting information signalled, reinforce our idea that this approach can be applied for other task of NLP.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

Towards the Automatic Merging of Lexical Resources: Automatic Mapping

  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Necşulescu, Silvia
Lexical Resources are a critical component for Natural Language Processing applications. However, the high cost of comparing and merging different resources has been a bottleneck to have richer resources with a broad range of potential uses for a significant number of languages.With the objective of reducing cost by/neliminating human intervention, we present a new method for automating the merging of resources,with special emphasis in what we call the mapping step. This mapping step, which converts the resources into a common format that allows latter the merging, is usually performed with huge manual effort and thus makes the whole process very costly. Thus, we propose a method to perform this mapping fully automatically. To test our method, we have addressed the merging of two verb subcategorization frame lexica for Spanish, The results/nachieved, that almost replicate human work, demonstrate the feasibility of the approach.
Repositorio Digital de la UPF
Publicación Contribución de congreso (ConferenceOutput)

A Method Towards the Fully Automatic Merging of Lexical Resources

  • Bel Rafecas, Núria
  • Padró, Muntsa
  • Necşulescu, Silvia
Lexical Resources are a critical component for Natural Language Processing applications. However, the high cost of comparing and merging different resources has been a bottleneck to obtain richer resources and a broader range of potential uses for a significant number of languages. With the objective of reducing cost by eliminating human intervention, we present a new method towards the automatic merging of resources. This method includes both, the automatic mapping of resources involved to a common format and merging them, once in this format. This paper presents how we have addressed the merging of two verb subcategorization frame lexica for Spanish, but our method will be extended to cover other types of Lexical Resources. The achieved results, that almost replicate human work, demonstrate the feasibility of the approach.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Towards Using Web-Crawled Data for Domain Adaptation in Statistical Machine Translation

  • Pecina, Pavel
  • Toral, Antonio
  • Way, Andy
  • Papavassiliou, Vassilis
  • Prokopidis, Prokopis
  • Giagkou, Maria
  • Toral, Antonio|||/items/ffa766b5-dbe2-4105-a0fc-e51f29946693
This paper reports on the ongoing work focused on domain adaptation of statistical machine translation using domain-specific data obtained by domain-focused web crawling. We present a strategy for crawling monolingual and parallel data and their exploitation for testing, language modelling, and system tuning in a phrase-based machine translation framework. The proposed approach is evaluated on the domains of Natural Environment and Labour Legislation and two language pairs: English–French and English–Greek.
RUA. Repositorio Institucional de la Universidad de Alicante
Publicación Otros (Other)

Simple and Effective Parameter Tuning for Domain Adaptation of Statistical Machine Translation

  • Pecina, Pavel
  • Toral, Antonio
  • van Genabith, Josef
  • Toral, Antonio|||/items/ffa766b5-dbe2-4105-a0fc-e51f29946693
Current state-of-the-art Statistical Machine Translation systems are based on log-linear models that combine a set of feature functions to score translation hypotheses during decoding. The models are parametrized by a vector of weights usually optimized on a set of sentences and their reference translations, called development data. In this paper, we explore a (common and industry relevant) scenario where a system trained and tuned on general domain data needs to be adapted to a specific domain for which no or only very limited in-domain bilingual data is available. It turns out that such systems can be adapted successfully by re-tuning model parameters using surprisingly small amounts of parallel in-domain data, by cross-tuning or no tuning at all. We show in detail how and why this is effective, compare the approaches and effort involved. We also study the effect of system hyperparameters (such as maximum phrase length and development data size) and their optimal values in this scenario.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Bilingual Dictionary Extraction Tools

  • Graham, Yvette
  • Thurmair, Gregor
  • Toral, Antonio
  • Aleksic, Vera
This document describes methods developed within task WP5.2. to automatically extract bilingual dictionaries from statistical machine translation (SMT) phrase tables, that were automatically generated from sentence aligned bilingual corpora (Deliverable 5.3).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

A Web Service-Based Platform for the Automatic Production of Language Resources

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Sample of Transfer Entries produced

Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

PANACEA: The Platform

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Architecture and design of the platform

  • Poch, Marc
  • Prokopidis, Prokopis
  • Thurmair, Gregor
  • Schnober, Carsten
  • Del Gratta, Riccardo
  • Bel Rafecas, Núria
  • Hamon, Olivier
This document aims to establish the requirements and the technological basis and design of the PANACEA platform. These are the main goals of the document: a) Survey the different technological approaches that can be used in PANACEA, b) Specify some guidelines for the metadata, c) Establish the requirements for the platform, d) Make a Common Interface proposal for the tools, e) Propose a format for the data to be exchanged by the tools (Travelling Object), f) Choose the technologies that will be used to develop the platform, and g) Propose a workplan.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Towards the automatic merging of language resources

  • Necşulescu, Silvia
  • Bel Rafecas, Núria
  • Padró, Muntsa
  • Marimon, Montserrat
  • Revilla, Eva
Language Resources are a critical component for Natural Language Processing applications. Throughout the years many resources were manually created for the same task, but with different granularity and coverage information. To create richer resources for a broad range of potential reuses, nformation from all resources has to be joined into one. The hight cost of comparing and merging different resources by hand has been a bottleneck for merging existing resources. With the objective of reducing human intervention, we present a new method for automating merging resources. We have addressed the merging of two verbs subcategorization frame (SCF) lexica for Spanish. The results achieved, a new lexicon with enriched information and conflicting information signalled, reinforce our idea that this approach can be applied for other task of NLP.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Finding dependency parsing limits over a large Spanish Corpus

  • Padró, Muntsa
  • Ballesteros, Miguel
  • Martínez, Héctor
  • Bohnet, Bernd
This paper studies the performance of different parsers over a large Spanish treebank. The aim of this work is to assess the limitations of state-of-the-art parsers. We want to select the most appropriate parser for subcategorization Frame acquisition, and we focus our analysis on two aspects: the accuracy drop when parsing out-of-domain data, and the performance over specific labels relevant to our task., HM has been partially funded by the European Commission’s 7th Framework Program under grant agreement 238405 (CLARA). MP has been funded by the European Project PANACEA (FP7-ICT-2010- 248064).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Tradução online

  • Poch, Marc
  • Bel Rafecas, Núria
  • Toral, Antonio
  • Hamon, Olivier
  • Quochi, Valeria
Um desafio estratégico no mundo globalizado atual é superar a barreira da língua através da tecnologia. Na Europa, um projeto está avaliando e desenvolvendo uma plataforma que poderá traduzir a enorme quantidade de dados (orais e escritos) produzidos pelas pessoas, num processo chamado "tradução online". Conheça a seguir os detalhes do estudo.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Merged dictionaries

  • Rimell, Laura
  • Bel Rafecas, Núria
  • Padró, Muntsa
  • Frontini, Francesca
  • Monachini, Monica
  • Quochi, Valeria
  • Del Gratta, Riccardo
This document presents the merged dictionaries delivered in PANACEA. Those dictionaries result from merging already existing lexica, generally for general domain, with domain specific lexica acquired using PANACEA platform. The domain specific lexica are presented and delivered in D6.3 and the merging repository that allowed the multilevel merging in D6.4.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Towards the Automatic Merging of Lexical Resources: Automatic Mapping

  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Necşulescu, Silvia
Lexical Resources are a critical component for Natural Language Processing applications. However, the high cost of comparing and merging different resources has been a bottleneck to have richer resources with a broad range of potential uses for a significant number of languages.With the objective of reducing cost by/neliminating human intervention, we present a new method for automating the merging of resources,with special emphasis in what we call the mapping step. This mapping step, which converts the resources into a common format that allows latter the merging, is usually performed with huge manual effort and thus makes the whole process very costly. Thus, we propose a method to perform this mapping fully automatically. To test our method, we have addressed the merging of two verb subcategorization frame lexica for Spanish, The results/nachieved, that almost replicate human work, demonstrate the feasibility of the approach.
Recercat. Dipósit de la Recerca de Catalunya
Publicación

Documentation of P clue/ lexical class from Weka computer Web Service

Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

PANACEA: Language Resource Factory. Data Available for free!

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Final Report on the Corpus Acquisition & Annotation subsystem and its components

  • Prokopidis, Prokopis
  • Papavassiliou, Vassilis
  • Toral, Antonio
  • Poch, Marc
  • Frontini, Francesca
  • Rubino, Francesco
  • Thurmair, Gregor
PANACEA WP4 targets the creation of a Corpus Acquisition and Annotation (CAA) subsystem for the acquisition and processing of monolingual and bilingual language resources (LRs). The/nCAA subsystem consists of tools that have been integrated as web services in the PANACEA platform of LR production. D4.2 Initial functional prototype and documentation in T13 and D4.4 Report on the revised Corpus Acquisition & Annotation subsystem and its components in T23 provided initial and updated documentation on this subsystem, while this deliverable presents the final documentation of the subsystem as it evolved after the third development cycle of the project. The deliverable is structured as follows. The Corpus Acquisition Component (i.e. the Focused Monolingual and Bilingual Crawlers (FMC/FBC)) is described in section 2. The final list of/ntools for corpus normalization (cleaning and de-duplication) is detailed in section 3. Section 4 provides documentation on all NLP tools included in the subsystem. Due to its nature, this deliverable aggregates considerable parts of all previous WP4 deliverables. The main new additions include a) new functionalities for, among others, crawling strategy, de-duplication, and detection of parallel document pairs; and b) new NLP tools for syntactic analysis, named entity recognition, tweet processing and anonymization.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Parallel technology tools and resources

  • Pecina, Pavel
  • Toral, Antonio
  • Thurmair, Gregor
  • Way, Andy
  • Schnober, Carsten
This report defines WP5 “Parallel corpus and derivatives” tasks in Section 3, describes the current state of the art in the relevant research areas in Section 4, provides analysis of tools that are available and specifies the tools and resources to be developed in this WP in Section 5 and Section 6, respectively, and proposes the solution path to be followed during the rest of the lifetime of the WP in Section 7.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Integrated Final Version of the Components for Lexical Acquisition

  • Rimell, Laura
  • Bel Rafecas, Núria
  • Padró, Muntsa
  • Frontini, Francesca
  • Monachini, Monica
  • Quochi, Valeria
This deliverable, D6.2, describes the development of lexical acquisition components for PANACEA, including the tools and technologies for each lexical acquisition task, and the integration of an appropriate subset of these tools into the PANACEA platform. In particular, the 3rd version of the platform has been released and represents the key time for lexical acquisition components to be deployed within PANACEA.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Third version (v4) of the integrated platform and documentation

  • Poch, Marc
  • Hamon, Olivier
  • Quochi, Valeria
  • Del Gratta, Riccardo
  • Toral, Antonio
  • Thurmair, Gregor
  • Prokopidis, Prokopis
  • Bel Rafecas, Núria
The 3rd version of the platform is working and WP4, WP5 and WP6 tools are deployed as web services. Common Interfaces (CI) and Travelling Object (TO) for new WP5 and WP6 have been devised.The Registry, deployed for the 1st version of the platform, is operational and has been updated with some new features. It has now more than 120 registered web services. The PANACEA/nmyExperiment portal was deployed to share workflows among users who can then execute those workflows with Taverna. Massive data solutions were developed during the second development cycles that have been used since then. New statistics, storage and security features have been addressed and studied. This deliverable will present all the work developed for the third version of the platform and its documentation.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

A Method Towards the Fully Automatic Merging of Lexical Resources

  • Bel Rafecas, Núria
  • Padró, Muntsa
  • Necşulescu, Silvia
Lexical Resources are a critical component for Natural Language Processing applications. However, the high cost of comparing and merging different resources has been a bottleneck to obtain richer resources and a broader range of potential uses for a significant number of languages. With the objective of reducing cost by eliminating human intervention, we present a new method towards the automatic merging of resources. This method includes both, the automatic mapping of resources involved to a common format and merging them, once in this format. This paper presents how we have addressed the merging of two verb subcategorization frame lexica for Spanish, but our method will be extended to cover other types of Lexical Resources. The achieved results, that almost replicate human work, demonstrate the feasibility of the approach.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Second evaluation report. Evaluation of PANACEA v2 and produced resources

  • Aleksic, Vera
  • Hamon, Olivier
  • Papavassiliou, Vassilis
  • Pecina, Pavel
  • Poch, Marc
  • Prokopidis, Prokopis
  • Quochi, Valeria
  • Schwartz, Christoph
  • Thurmair, Gregor
This deliverable reports on the second evaluation cycle consisting of: 1) the validation of the platform v2, i.e. the integration of components; and 2) the evaluation of the components that produce resources, and, therefore, of the resources produced. The methodology and criteria for the evaluation of the technology integrated into the platform and for the validation of the integration of components have been described in D7.1. Some of the criteria involved in this evaluation cycle will be repeated here for/nthe reader’s sake.
Recercat. Dipósit de la Recerca de Catalunya
Publicación

[PANACEA] Project final report

  • Bel Rafecas, Núria
This report describes the results of PANACEA project. PANACEA project has focused on the development of a factory of LRs that automates the stages involved in/nthe acquisition, production, updating and maintenance of LRs required by MT systems, and by other based on Language Technologies (LT) applications. This automation is meant to cut down costs significantly, in terms of time and human effort. Such reductions are the only way to guarantee a continuous supply of LRs that MT/nand other Language Technologies may demand in a multilingual Europe. In order to address this objective, PANACEA has worked in (i) the development of a platform, designed as a dedicated factory for the composition of a number of LRs production lines based on combinations of different web services and (ii) the integration of advanced components for the acquisition and normalization of corpora, monolingual and parallel corpora, their alignment; the derivation of bilingual dictionaries out of aligned corpora; and the production of monolingual rich information lexica using corpus based automatic methods.
Recercat. Dipósit de la Recerca de Catalunya
Publicación

Documentation of MaltParser Web Service

Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

[What is it?] PANACEA: A Web Service-Based Platform for the Automatic Production of Language Resources

Recercat. Dipósit de la Recerca de Catalunya
Publicación

Documentation of Nouns Classifier Web Service

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Verb SCF extraction for Spanish with dependency parsing = Extracción de patrones de subcategorización de verbos en castellano con análisis de dependencias

  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Garí Soler, Aina
En este artículo presentamos los resultados de nuestros experimentos en producción automática de léxicos con información de patrones de subcategorización verbal para castellano. La investigación se llevó a cabo en el marco del proyecto PANACEA de adquisición automática de información léxica que redujera al máximo la intervención humana. En nuestros experimentos, se utilizó una cadena de diferentes herramientas que incluía ‘crawling’ de textos de un dominio particular, normalización y limpieza de los textos, segmentación, identificación de unidades, etiquetado categorial y análisis de dependencias antes de, finalmente, la extracción de los patrones de subcategorización. Los resultados obtenidos muestran una gran dependencia de la calidad de los analizadores de dependencias aunque, no obstante, están en línea con los resultados obtenidos en experimentos similares para otras lenguas., In this paper we present the results of our experiments in automatic production of verb subcategorization frame lexica for Spanish. The work was carried out in the framework of the PANACEA project aiming at the automatic acquisition of lexical information reducing at maximum human intervention. In our experiments, a chain of different tools was used: domain focused web crawling, automatic cleaning, segmentation and tokenization, PoS tagging, dependency parsing and finally SCFs extraction. The obtained results show a high dependency on the quality of the results of the intervening components, in particular of the dependency parsing, which is the focus of this paper. Nevertheless, the results achieved are in line with the state-of-the-art for other languages in similar experiments., This work was funded by the European Project PANACEA (FP7-ICT-2010-248064).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Platform Software, Project Tools + Resources Licensing Policy and Exploitation Plan

  • Arranz, Victoria
  • Choukri, Khalid
  • Hamon, Olivier
  • Bel Rafecas, Núria
  • Tsiavos, Prodromos
This report elaborates on the exploitation of the PANACEA project assets. These assets have been clustered into a few items (a) the PANACEA Factory/Platform, (b) the web services integrated within the platform, (c) the associated workflows to manage the sequencing of web services (d) the tools developed during the project and last but not least (e) the data sets, i.e. Language Resources (LRs) produced mainly for Machine translation/localization, but not only, within the project, exploiting the platform, web services, and the workflows.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

PANACEA, Platform for Automatic, Normalized Annotation and Cost-Effective Acquisition of Language Resources for Human Language Technologies

Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Efficiency-based evaluation of aligners for industrial applications

  • Toral, Antonio
  • Poch, Marc
  • Pecina, Pavel
  • Thurmair, Gregor
This paper presents a novel efficiencybased evaluation of sentence and word aligners. This assessment is critical in order to make a reliable use in industrial scenarios. The evaluation shows that the resources/nrequired by aligners differ rather broadly. Subsequently, we establish limitation mechanisms on a set of aligners deployed as web services. These results, paired with the quality expected from the aligners, allow providers to choose the most appropriate aligner according to the task at hand.
Recercat. Dipósit de la Recerca de Catalunya
Publicación

Documentation of Bayesian parameter estimation Web Service

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Initial functional prototype and documentation describing the initial CAA subsystem and its components

  • Prokopidis, Prokopis
  • Papavassiliou, Vassilis
  • Toral, Antonio
  • Arranz, Victoria
  • Bel Rafecas, Núria
This document focuses on the development and integration of the first version of the Corpus Acquisition and Annotation (CAA)/nsubsystem in the PANACEA platform. This version incorporates a Corpus Acquisition Component (CAC) and a Cleanup and Normalization Component (CNC) as planned in Section 7 of D4.1 Technologies and tools for corpus creation, normalization and annotation. The present deliverable, together with D4.3, constitutes the second milestone of WP4.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Otros (Other)

PANACEA: PAROLE Association Workshop 18/19 Octobre 2012

Recercat. Dipósit de la Recerca de Catalunya
Publicación

Documentation of Naive Bayes classifier Web Service

Recercat. Dipósit de la Recerca de Catalunya
Publicación

Travelling Object definition for multilevel lexicon in PANACEA platform

  • Rimell, Laura
  • Bel Rafecas, Núria
  • Padró, Muntsa
  • Quochi, Valeria
  • Frontini, Francesca
  • Monachini, Monica
In this document, the format of the lexica acquired and delivered by PANACEA platform is specified. The targeted standard format chosen for these travelling objects is basic Lexical Markup Framework, LMF (Francopoulo et al. 2008). Below we provide LMF examples for the levels of concern in PANACEA, all of them based on LMF using DTD in revision 16.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

User’s Workshop: Technology transfer, papers produced and/ndissemination materials

  • Arranz, Victoria
  • Hamon, Olivier
  • Mazo, Hélene
  • Leixa, Jéremy
  • Bel Rafecas, Núria
The PANACEA user’s workshop took place in Budapest on June 29, as a satellite event of the META-FORUM 2011. The choice for this collocated event was indeed strategic. META-FORUM is an event organised by another FP7 project (T4ME/META-NET) that shares a strong interest for the industrial world with PANACEA. This seemed very interesting in order to attract potential users’ interest and encourage them to join us for the workshop. Targeting two events was certainly a plus for some of the attendees.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Transfer Selection Support

Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Mining and exploiting domain-specific corpora in the PANACEA platform

  • Bel Rafecas, Núria
  • Papavassiliou, Vassilis
  • Prokopidis, Prokopis
  • Toral, Antonio
  • Arranz, Victoria
The objective of the PANACEA ICT-2007.2.2 EU project is to build a platform that automates the stages involved in the acquisition,production, updating and maintenance of the large language resources required by, among others, MT systems. The development of a Corpus Acquisition Component (CAC) for extracting monolingual and bilingual data from the web is one of the most innovative building blocks of PANACEA. The CAC, which is the first stage in the PANACEA pipeline for building Language Resources, adopts an efficient and distributed methodology to crawl for web documents with rich textual content in specific languages and predefined domains. The CAC includes modules that can acquire parallel data from sites with in-domain content available in more than one language. In order to extrinsically evaluate the CAC methodology, we have conducted several experiments that used crawled parallel corpora for the identification and extraction of parallel sentences using sentence alignment. The corpora were then successfully used for domain adaptation of Machine Translation Systems.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Second version (v2) of the integrated platform and documentation

  • Poch, Marc
  • Hamon, Olivier
  • Toral, Antonio
  • Prokopidis, Prokopis
  • Bartolini, Roberto
  • Rubino, Francesco
  • Thurmair, Gregor
  • Papavassiliou, Vassilis
  • Bel Rafecas, Núria
The 2nd version of the platform is working and WP4 and WP5 aligners are deployed. The Registry, deployed for the 1st version of the platform, is operational and it has now 70 registered web services. The PANACEA myExperiment portal has been deployed to share workflows among users who can execute those workflows with Taverna. Massive data solutions have been developed for users and web service providers and workflows have been redesigned and improve to handle larger amounts of data. This deliverable will present all the work developed for WP3 and its documentation.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Analysis of Industrial User Requirements

Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Automatic Extraction of Polar Adjectives for the Creation of Polarity Lexicons

  • Vázquez, Silvia
  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Gonzalo, Julio
Automatic creation of polarity lexicons is a crucial issue to be solved in order to reduce time and/nefforts in the first steps of Sentiment Analysis. In this paper we present a methodology based on/nlinguistic cues that allows us to automatically discover, extract and label subjective adjectives/nthat should be collected in a domain-based polarity lexicon. For this purpose, we designed a/nbootstrapping algorithm that, from a small set of seed polar adjectives, is capable to iteratively/nidentify, extract and annotate positive and negative adjectives. Additionally, the method/nautomatically creates lists of highly subjective elements that change their prior polarity even/nwithin the same domain. The algorithm proposed reached a precision of 97.5% for positive/nadjectives and 71.4% for negative ones in the semantic orientation identification task.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Web Services for Bayesian Learning

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Criteria for evaluation of resources, technology and integration

  • Caselli, Tommaso
  • Quochi, Valeria
  • Arranz, Victoria
  • Bel Rafecas, Núria
  • Hamon, Olivier
  • Papavassiliou, Vassilis
  • Poch, Marc
  • Thurmair, Gregor
  • Toral, Antonio
  • Strik Lievers, Francesca
  • Rimell, Laura
This deliverable defines how evaluation will be carried out at each integration cycle. As PANACEA aims at producing large scale resources, evaluation becomes a critical and challenging issue. Critical because it is important to assess the quality of the results that should be delivered to users. Challenging because we/nprospect rather new areas, and through a technical platform: some new methodologies will have to be explored or old ones to be adapted.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Using qualia information to identify lexical semantic classes in an unsupervised clustering task

  • Romeo, Lauren
  • Mendes, Sara
  • Bel Rafecas, Núria
Acquiring lexical information is a complex problem, typically approached by relying on a number of contexts to contribute information for classification. One of the first issues to address in this domain is the determination of such contexts. The work presented here proposes the use of automatically obtained FORMAL role descriptors as features used to draw nouns from the same lexical semantic class together in an unsupervised clustering task. We have dealt with three lexical semantic classes (HUMAN, LOCATION and EVENT) in English. The results obtained show that it is possible to discriminate between elements from different lexical semantic classes using only FORMAL role information, hence validating our initial hypothesis. Also, iterating our method accurately accounts for fine-grained distinctions within lexical classes, namely distinctions involving ambiguous expressions. Moreover, a filtering and bootstrapping strategy employed in extracting FORMAL role descriptors proved to minimize effects of sparse data and noise in our task.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Towards a User-Friendly Platform for Building Language Resources based on Web Services

  • Poch, Marc
  • Toral, Antonio
  • Hamon, Olivier
  • Quochi, Valeria
  • Bel Rafecas, Núria
This paper presents the platform developed in the PANACEA project, a distributed factory that automates the stages involved in the acquisition, production, updating and maintenance of Language Resources required by Machine Translation and other Language Technologies. We adopt a set of tools that have been successfully used in the Bioinformatics field, they are adapted to the needs of our field and used to deploy web services, which can be combined to build more complex processing chains (workflows). This paper describes the platform and its different components (web services, registry, workflows, social network and interoperability). We demonstrate the scalability of the platform by carrying out a set of massive data experiments. Finally, a validation of the platform across a set of required criteria proves its usability for different types of users (non-technical users and providers).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Report on the revised Corpus Acquisition & Annotation subsystem and its components

  • Prokopidis, Prokopis
  • Papavassiliou, Vassilis
  • Toral, Antonio
  • Poch, Marc
  • Frontini, Francesca
  • Rubino, Francesco
  • Thurmair, Gregor
PANACEA WP4 targets the creation of a Corpus Acquisition and Annotation (CAA) subsystem for the acquisition and processing of monolingual and bilingual language resources (LRs). The CAA subsystem consists of tools that have been integrated as web services in the PANACEA platform of LR production. D4.2 Initial functional prototype and documentation in T13 provided documentation on the initial functional prototype of this subsystem, while this deliverable presents updates in the revised subsystem during the second development cycle of the project. The deliverable is structured as follows. A revised version of the Focused Monolingual Crawler (FMC), that has been implemented according to the results of the first evaluation cycle and the reviewers’ comments in the first annual review report, is described in section 2. New and revised versions of tools for corpus normalization (cleaning and deduplication) are detailed in section 3. Section 4 provides documentation on the NLP tools introduced for the first time in the subsystem. These tools focus mainly on sentence splitting/tokenization and POS tagging/lemmatization for English (EN), French (FR), Spanish (ES), German (DE), Italian (IT) and Greek (EL).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

PANACEA (Platform for Automatic, Normalised Annotation and Cost-Effective Acquisition of Language Resources for Human Language Technologies)

  • Bel Rafecas, Núria
  • Poch, Marc
  • Toral, Antonio
The objective of PANACEA is to build a factory of LRs that automates the stages involved in the acquisition, production, updating and maintenance of LRs required by MT systems and by other applications based on language technologies, and simplifies eventual issues regarding intellectual property rights. This automation will cut down the cost, time and human effort significantly. These reductions of costs and time are the only way to guarantee the continuous supply of LRs that MT and other language technologies will be demanding in the multilingual Europe.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Technologies and tools for corpus creation, normalization and annotation

  • Prokopidis, Prokopis
  • Papavassiliou, Vassilis
  • Pecina, Pavel
  • Rimell, Laura
  • Poibeau, Thierry
  • Bartolini, Roberto
  • Caselli, Tommaso
  • Frontini, Francesca
  • Aleksic, Vera
  • Thurmair, Gregor
  • Poch, Marc
  • Bel Rafecas, Núria
  • Hamon, Olivier
The objectives of the Corpus Acquisition and Annotation (CAA) subsystem are the acquisition and processing of monolingual and bilingual language resources (LRs) required in the PANACEA context. Therefore, the CAA subsystem includes: i) a Corpus Acquisition Component (CAC) for extracting monolingual and bilingual data from the web, ii) a component for cleanup and normalization (CNC) of these data and iii) a text processing component (TPC) which consists of NLP tools including modules for sentence splitting, POS tagging, lemmatization, parsing and named entity recognition. This report presents the terminology used in this document in Section 2. The report explaines state-of-the-art and existing tools for corpus acquisition, corpus normalization, and text processing in Sections 3, 4 and 5 respectively. The resources to be produced in the context of WP4 are discussed in Section/n6. In Section 7 it presents the solution path we aim to explore for generating these resources.
Recercat. Dipósit de la Recerca de Catalunya
Publicación

Documentation of Select Nouns from LMF lexicon Web Service

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Aligners Integrated Into The Platform

  • Toral, Antonio
  • Pecina, Pavel
  • Way, Andy
This document constitutes the second milestone of work-package 5. It reports on the integration into the Panacea Platform of the aligners described in Deliverable 5.1 (Section 5 and appendix A).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Task-based Evaluation of the PANACEA Production Chain

  • Aleksic, Vera
  • Schwartz, Christoph
  • Thurmair, Gregor
  • Prokopidis, Prokopis
  • Papavassiliou, Vassilis
  • Toral, Antonio
The work package 8 of PANACEA deals with ‘Industrial Evaluation’; after three development cycles, the final task is to relate the PANACEA developments to real world scenarios. The package has two/ntasks: a) Tool-based evaluation looks at the main tools, and evaluates their usability in industrial contexts; this is subject of the deliverable D8.2; and b) Task-based evaluation uses PANACEA tools to perform a given task, and compares the outcome with conventional procedures. This is subject of the current deliverable (D8.3).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Third evaluation report. Evaluation of PANACEA v3 and produced resources

  • Quochi, Valeria
  • Frontini, Francesca
  • Bartolini, Roberto
  • Hamon, Olivier
  • Poch, Marc
  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Thurmair, Gregor
  • Toral, Antonio
  • Kamram, Amir
D7.4 reports on the evaluation of the different components integrated in the PANACEA third cycle of development as well as the final validation of the platform itself. All validation and evaluation experiments follow the evaluation criteria already described in D7.1. The main goal of WP7 tasks was to test the (technical) functionalities and capabilities of the middleware that allows the integration of the various resource-creation components into an interoperable distributed environment (WP3) and to/nevaluate the quality of the components developed in WP5 and WP6. The content of this deliverable is thus complementary to D8.2 and D8.3 that tackle advantages and usability in industrial scenarios. It/nhas to be noted that the PANACEA third cycle of development addressed many components that are still under research. The main goal for this evaluation cycle thus is to assess the methods experimented with and their potentials for becoming actual production tools to be exploited outside research labs.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Lexical Merger

  • Del Gratta, Riccardo
  • Monachini, Monica
  • Tesconi, Maurizio
  • Abrate, Matteo
  • Lo Duca, Angelica
  • Rimell, Laura
  • Bel Rafecas, Núria
  • Padró, Muntsa
This document describes the experiments on the merging of lexical resources performed during the project and the development of two merging components for LMF lexicons.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Interoperability and Technology for a Language Resources Factory

  • Poch, Marc
  • Bel Rafecas, Núria
This document describes some of the technological aspects of a project devoted to the creation of a factory for language resources. The project’s objectives are explained, as well as the idea to create a distributed infrastructure of web services. This document focuses on two main topics of the factory: (1) the technological approaches chosen to develop the factory, i.e. software, protocols, servers, etc. (2) and Interoperability as the main challenge is to permit different NLP tools work together in the factory. This document explains why XCES and GrAF are chosen as the main formats used for the linguistic data exchange., This research has been funded by the PANACEA project (EU-7FP-ITC-248064).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Monolingual lexica for English, Spanish and Italian tuned for a particular domain (LAB and ENV)

  • Rimell, Laura
  • Bel Rafecas, Núria
  • Padró, Muntsa
  • Frontini, Francesca
  • Monachini, Monica
  • Quochi, Valeria
  • Del Gratta, Riccardo
This document presents the lexica acquired using PANACEA platform for Labour and Environment domains and delivered in D6.3. The languages of the lexica are English, Spanish and Italian. The lexical information acquired depends on the language, according to the available tools in the platform. Next section lists the produced lexica and the languages for which they are available.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

A Classification of Adjectives for Polarity Lexicons Enhancement

  • Vázquez, Silvia
  • Bel Rafecas, Núria
Subjective language detection is one of the most important challenges in Sentiment Analysis. Because of the weight and frequency in opinionated texts, adjectives are considered a key piece in the opinion extraction process. These subjective units are more and more frequently collected in polarity lexicons in which they appear annotated with their prior polarity. However, at the moment, any polarity lexicon takes into account prior polarity variations across domains. This paper proves that a majority of adjectives change their prior polarity value depending on the domain. We propose a distinction between domain dependent and romain independent adjectives. Moreover, our analysis led us to propose a further classification related to subjectivity degree: constant, mixed and highly subjective adjectives. Following this classification, polarity values will be a better support for Sentiment Analysis.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Towards the Fully Automatic Merging of Lexical Resources: A Step Forward

  • Padró, Muntsa
  • Bel Rafecas, Núria
  • Necşulescu, Silvia
This article reports on the results of the research done towards the fully automatically merging of lexical resources. Our main goal is to show the generality of the proposed approach, which have been previously applied to merge Spanish Subcategorization Frames lexica. In this work we extend and apply the same technique to perform the merging of morphosyntactic lexica encoded in LMF. The experiments showed that the technique is general enough to obtain good results in these two different tasks which is an important step towards performing the merging of lexical resources fully automatically.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

English-French and English-Greek bilingual dictionaries for the Environment and Labour Legislation domains

  • Graham, Yvette
  • Thurmair, Gregor
  • Toral, Antonio
  • Aleksic, Vera
This report describes the deliverable D5.5 of the PANACEA project: Greek-English and French-English bilingual dictionaries for the environment and legislative domains automatically extracted from parallel corpora. This deliverable is one of the outcomes of the task WP5.2. The resulting domain-specific dictionaries include a total of 80,000 bilingual entries for two language pairs:English–French and English–Greek and two domains: Environment and Labour Legislation.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Otros (Other)

PANACEA: LREC 2012, EC Projects Session

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Technologies and Tools for Lexical Acquisition

  • Rimell, Laura
  • Korhonen, Anna
  • Quochi, Valeria
  • Bel Rafecas, Núria
  • Caselli, Tommaso
  • Prokopidis, Prokopis
  • Gavrilidou, Maria
  • Poibeau, Thierry
  • Padró, Muntsa
  • Revilla, Eva
  • Monachini, Monica
  • Tesconi, Maurizio
  • Abrate, Matteo
  • Bacciu, Clara
This report describes the technologies and tools to be used for Lexical Acquisition in PANACEA. It includes descriptions of existing technologies and tools which can be built on and improved within PANACEA, as well as of new technologies and tools to be developed and integrated in PANACEA. The report touches briefly on the criteria for evaluating the results of the tools and integration (but see D7.1 for a detailed description of the evaluation criteria). It also specifies the Lexical Resources to be produced.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Tool-based Evaluation of the PANACEA Production Chain

  • Aleksic, Vera
  • Schwartz, Christoph
  • Thurmair, Gregor
  • Prokopidis, Prokopis
  • Papavassiliou, Vassilis
  • Toral, Antonio
After the end of the third evaluation cycle, the PANACEA project launched a final work package, dealing with the evaluation of the PANACEA results in industrial contexts. One focus was the usability of the tools in larger practical setups. A given domain (health & safety) and a given language direction (Italian to German) were investigated, and all tools involved from crawling to bilingual/nlexicon generation were applied. The main evaluation criterion was usability in industrial contexts; this comprises robustness, effort,/nand achieved quality of the single tools in comparison with other (state-of-the-art) tools; detailed evaluation was already done in the development cycles, and should not be repeated here.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

First version (v1) of the integrated platform/nand documentation

  • Poch, Marc
  • Hamon, Olivier
  • Thurmair, Gregor
  • Bel Rafecas, Núria
The aim of this document is to list and explain the work developed in WP3 "The platform" and the resulting documentation.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

First evaluation report. Evaluation of PANACEA v1 and produced resources

  • Arranz, Victoria
  • Caselli, Tommaso
  • Hamon, Olivier
  • Papavassiliou, Vassilis
  • Pecina, Pavel
  • Poch, Marc
  • Prokopidis, Prokopis
  • Quochi, Valeria
This deliverable reports on the first evaluation cycle consisting of: 1) the validation of the platform, i.e. the integration of components; and 2) the evaluation of the components that produce resources, and, therefore, of the resources produced. The methodology and criteria for the evaluation of the technology integrated into the platform and for the validation of the integration of components have been described in D7.1. Some of the/ncriteria involved in this evaluation cycle will be repeated here for the reader’s sake.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

PANACEA: A Web Service-Based Platform for the Automatic Production of Language Resources

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Platform for Automatic, Normalized Annotation and Cost-Effective Acquisition of Language Resources for Human Language Technologies: PANACEA

  • Bel Rafecas, Núria
El objetivo de PANACEA es engranar diferentes herramientas avanzadas para construir una fábrica de Recursos Lingüísticos (RL), una línea de producción que automatice los pasos implicados en la adquisición, producción, actualización y mantenimiento de los RL que la Traducción Automática y otras tecnologías lingüísticas, necesitan.
Recercat. Dipósit de la Recerca de Catalunya
Publicación

Documentation of LMF file merger Web Service

Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

PANACEA: Language Technology Days. Luxembourg, 22/23 March 2010

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

English-French and English-Greek parallel corpus for the Environment and Labour Legislation domains

  • Pecina, Pavel
  • Toral, Antonio
  • Papavassiliou, Vassilis
  • Prokopidis, Prokopis
  • Arranz, Victoria
This report describes the deliverable D5.3 of the PANACEA project: parallel, sententially aligned texts, cleaned and prepared for training-building translation models. This deliverable is/nan outcome of the work packages WP4.1, WP4.2, and WP5.1. The resulting domain-specific parallel corpus includes a total of 4 million words for two language pairs: English–French and/nEnglish–Greek and two domains: Environment and Labour Legislation. The data for each language pair and domain is split into three sets: training data for building translation models, development data for development purposes, and test data for testing and evaluation.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Language Resources Factory: case study on the acquisition of Translation Memories

  • Poch, Marc
  • Toral, Antonio
  • Bel Rafecas, Núria
This paper demonstrates a novel distributed architecture to facilitate the acquisition of Language Resources. We build a factory that automates the stages involved in the acquisition, production, updating and maintenance of these resources. The factory is designed as a platform where functionalities are deployed as web services, which can be combined in complex acquisition chains using workflows. We show a case study, which acquires a Translation Memory for a given pair of languages and a domain using web services for crawling, sentence alignment and conversion to TMX.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Automatic Lexical Semantic Classification of Nouns

  • Bel Rafecas, Núria
  • Romeo, Lauren
  • Padró, Muntsa
The work we present here addresses cue-based noun classification in English and Spanish. Its main objective is to automatically acquire lexical semantic information by classifying nouns into previously known noun lexical classes. This is achieved by using particular aspects of linguistic contexts as cues that identify a specific lexical class. Here we concentrate on the task of identifying such cues and the theoretical background that allows for an assessment of the complexity of the task. The results show that, despite of the a-priori complexity of the task, cue-based classification is a useful tool in the automatic acquisition of lexical semantic classes.
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

Monolingual corpus acquired in five languages and two domains

  • Papavassiliou, Vassilis
  • Prokopidis, Prokopis
  • Toral, Antonio
  • Arranz, Victoria
  • Bel Rafecas, Núria
  • Quochi, Valeria
This deliverable document explains the creation process of monolingual corpora that were created with the first version of the Corpus Acquisition and Annotation subsystem (CAA). The first version of the CAA was developed during the first year of the project following insights from D4.1 Technologies and tools for corpus creation, normalization and annotation (T6). The main component currently integrated in the CAA is the Corpus Acquisition Component (CAC). Although the CAC is described in more detail in D4.2 Initial functional prototype and documentation (due T13), we include extracts from that document in the present deliverable, when we believe that this would help the reader understand the corpus acquisition process we followed.
Recercat. Dipósit de la Recerca de Catalunya
Publicación

Documentation of Weka noun signatures creator Web Service

Recercat. Dipósit de la Recerca de Catalunya
Publicación Contribución de congreso (ConferenceOutput)

Automatic Detection of Non-deverbal Event Nouns for Quick Lexicon Production

  • Bel Rafecas, Núria
  • Coll, Maria
  • Resnik, Gabriela
In this work we present the results of experimental work on the development of lexical class-based lexica by automatic means. Our purpose is to assess the use of linguistic lexical-class based information as a feature selection methodology for the use of classifiers in quick lexical development. The results show that the approach can help reduce the human effort required in the development of language resources significantly.