Search

Poster presentation for HDLS14 with poster PDF, videorecorded presentation, subtitles and image descriptions

In this article, we present Procura-PALavras (P-PAL), a Web-based interface for a new European Portuguese (EP) lexical database. Based on a contemporary printed corpus of over 227 million words, P-PAL provides a broad range of word attributes and statistics, including several measures of word frequency (e.g., raw counts, per-million word frequency, logarithmic Zipf scale), morpho-syntactic information (e.g., parts of speech [PoSs], grammatical gender and number, dominant PoS, and frequency and relative frequency of the dominant PoS), as well as several lexical and sublexical orthographic (e.g., number of letters; consonant-vowel orthographic structure; density and frequency of orthographic neighbors; orthographic Levenshtein distance; orthographic uniqueness point; orthographic syllabification; and trigram, bigram, and letter type and token frequencies), and phonological measures (e.g., pronunciation, number of phonemes, stress, density and frequency of phonological neighbors, transposed and phonographic neighbors, syllabification, and biphone and phone type and token frequencies) for 53,000 lemmatized and 208,000 nonlemmatized EP word forms. To obtain these metrics, researchers can choose between two word queries in the application: (i) analyze words previously selected for specific attributes and/or lexical and sublexical characteristics, or (ii) generate word lists that meet word requirements defined by the user in the menu of analyses. For the measures it provides and the flexibility it allows, P-PAL will be a key resource to support research in all cognitive areas that use EP verbal stimuli. P-PAL is freely available at http://p-pal.di.uminho.pt/tools., This study was conducted at the Psychology Research Centre (PSI/01662), University of Minho. It was supported by the Portuguese Foundation for Science and Technology and the Portuguese Ministry of Science, Technology, and Higher Education through national funds, and co-financed by FEDER through European funds (Grant POCI-01-0145-FEDER-007653).

Iako vrlo jednostavne na prvi pogled, riječi su izrazito kompleksne naravi, što je brojne istraživače posljednjih godina potaknulo na znanstvena istraživanja u kojima nastoje objasniti zakonitosti po kojima se riječi usvajaju i obrađuju. Poznato je kako mnogobrojni jezični i izvanjezični faktori utječu na pohranu i procesiranje jezičnih informacija. Čimbenici koji su odgovorni za brže i točnije prizivanje riječi iz mentalnog leksikona u jezičnoj percepciji, razumijevanju i produkciji izjednačavaju se s pojmom psiholingvističkih varijabli, a aktualna su tema u području psiholingvističkih istraživanja. Cilj ovog rada je pružiti jasan pregled psiholingvističkih obilježja riječi poput predočivosti, čestotnosti, apstraktnosti i konkretnosti, dobi usvajanja riječi i ostalih obilježja te pokazati kako ona utječu na leksičku obradu. Svrha ovog rada je, dakle, prikazati neke dosadašnje spoznaje i aktualne projekte iz relevantne znanstvene i stručne literature o psiholingvističkim čimbenicima koji utječu na leksičku obradu. U radu je prikazano nekoliko relevantnih modela jezične obrade, koji dominiraju stručnom literaturom u objašnjenju leksičkog pristupa, obrade i proizvodnje. Nadalje, prikazane su i recentne spoznaje o neuroanatomskoj pozadini jezičnih funkcija, razina i zadataka. Utvrđeno je kako su konkretnost, čestotnost, dob usvajanja, predočivost i poznatost riječi važni čimbenici koji značajno utječu na brzinu i točnost jezičnog procesiranja i kod jednojezičnih i dvojezičnih govornika, kao i kod govornika s jezičnim teškoćama. Zaključuje se kako suvremena literatura ukazuje na važnost psiholingvističkih čimbenika, objavljujući znanstvene radove i razvijajući javno dostupne leksičke baze te da je navedena obilježja riječi nužno kontrolirati u istraživačke i kliničke svrhe, budući da značajno utječu na jezično procesiranje., Although very simple at first, words are extremely complex, which has prompted many researchers in recent years to develop scientific studies in which they try to explain the laws by which words are adopted and processed. It is wll-known that many linguistic and extralinguistic factors affect the storage and processing of linguistic information. Those factors responsible for faster and more accurate word retrieval from the mental lexicon in linguistic perception, understanding and production are equated with the notion of psycholinguistic variables and are a current topic in the field of psycholinguistic research. The aim of this paper is to provide a clear overview of psycholinguistic factors such as word imageability, frequency, abstractness and concreteness, age of acquisition and other factors, and to show how they affect lexical processing. The purpose of this paper is, therefore, to present some current knowledge and current projects from the relevant scientific literature of psycholinguistic factors that affect lexical processing. The paper presents several relevant models of language processing, which dominate the literature in explaining the lexical approach, processing, and production. Furthermore, recent insights into the neuroanatomical background of language functions, levels, and tasks are presented. The word concretness, frequency, age of aquisition, imageability and familiarity have been found to be important factors that significantly affect the speed and accuracy of language processing in both monolingual and bilingual speakers, as well as in speakers with language difficulties. It is concluded that modern literature indicates the importance of psycholinguistic factors, publishing scientific papers and developing publicly available lexical databases, and that these word characteristics must be controlled for research and clinical purposes, as they significantly affect language processing.

Neste trabalho apresentamos as estratégias e os procedimentos adoptados na constituição de uma nova medida de frequência lexical do Português Europeu contemporâneo, o Procura-PALavras (P-PAL). Baseado num corpus de mais de 227 milhões de palavras, o P-PAL é uma aplicação web que oferece, por defeito, valores de frequência lexical para todas as suas entradas lexicais (lemas e formas) e que permite a computação de uma grande diversidade de outras medidas objectivas (lexicais e sublexicais) e subjectivas. Descrevemos ainda o processo de definição das suas entradas lexicais e da extracção dos respectivos valores de frequência. O elevado número de índices e de entradas lexicais tornam o P-PAL numa aplicação web avançada e indispensável à promoção e internacionalização da investigação em Portugal. O P-PAL encontra-se disponível em http://p-pal.di.uminho.pt/tools In this paper we present the strategies and procedures undertaken in the development of a new measure of lexical frequency of the contemporary European Portuguese - Procura-PALavras (P-PAL). Based on a corpus of over 227 million words, P-PAL offers the default frequency per million words (lemmas and wordforms), and the computation of several other objective (lexical and sublexical) and subjective word metrics. We also describe lexical entry integration and word frequency extraction. The high number of indices and lexical entries makes P-PAL an advanced and indispensable web application for the promotion and internationalization of Portuguese research. P-PAL is available at http://p-pal.di.uminho.pt/tools

The presentation of research results concerning applicability of AOO and EWN in machine translation is divided into two parts. In the first part, the author presents the influence of databases design, their theoretical aspects and the categorization of lexical items on the process and result of machine translation. The second part is devoted to the internal hierarchy of databases including the established categorization of lexical items as well as semantic inheritance comprising multiple inheritance and disambiguation accomplished by the above mentioned databases. The presented results coupled with the capability of contemporary computer technology give grounds for claiming that a bilingual approach in lexical description for machine translation is the best choice.

Cataloged from PDF version of article. We present an efficient algorithm for retrieving from a database of trees, all trees that differ from a given query tree by a small number additional or missing leaves, or leaf label changes. It has natural language processing applications in searching for matches in example-based translation systems, and retrieval from lexical databases containing entries of complex feature structures. For large randomly generated synthetic tree databases (some having tens of thousands of trees), and on databases constructed from Wall Street Journal treebank, it can retrieve for trees with a small error, in a matter of tenths of a second to about a second.

Lexical databases such as AOO or EWN are built in order to enable machine translation of written word. The author’s intention is to test the applicability of the above listed databases in machine translation and to choose the one that proves to be more successful. The presentation of final results consists of two parts. Part I is the analysis of factors influencing the process of machine translation with the use of AOO and EWN that is to say: steps in database design, their theoretical aspects and the categorization of lexical items. In author’s belief, the above mentioned elements exert profound influence on the result of machine translation produced with the use of one of the herein described lexical databases. The second part of the presentation touches on the matters of hierarchy, semantic inheritance and word-sense disambiguation.

This paper describes an ongoing research conducted at the Institute for Computational Linguistics (ILC) of the National Research Council (CNR) in Pisa, concerning the creation of a lexical semantic database for the Latin language, with its Italian and English translation. As we all know, many words of the Italian language derive from Latin, in the same way as many Latin words and expressions are used in everyday Italian language. The aim of this work is to exploit a well-known and tested conceptual model to construct a resource that can be easily consulted by the students of high school and university who wish to achieve proficiency in Latin language learning and deeper awareness of the use of Italian. The database is of a relational type, following the EuroWordNet (EWN)/ItalWordNet (IWN) model, according to the WordNet (WN) philosophy perspective. The items included in the database are clustered into approximately 250 synsets, which are sets of one or more synonyms, for example carina, navis (ship). Each Latin synset is linked to other words (or synsets) by means of semantic 'internal relations' of the hierarchical-vertical type (or 'is-a' relations) and of the horizontal type (role, means, purpose, sub-event, instance, etc. relations), on the basis of the model. One of the most important characteristics of the database is the possibility of connecting the Latin words to the equivalent English synonyms (or near synonyms), included in the Princeton database WordNet, by means of 'equivalence relations'. 'Plug-in relations connect each word of the Latin lexicon to the equivalent synset (synonym) of the Italian WordNet (IWN), as a kind of hookup point from which it is possible to see the Italian synonym with all its semantic relations. The study of all these kinds of relations can support the students in the learning of the Latin language, linked to Italian as well as to English. The database currently contains 200 nouns and 50 verbs, drawn from the most recent version of the Campanini-Carboni's Latin and Italian Dictionary (2011), and from William's Vocabulary for Speaking Latin (1829). A set of adjectives and adverbs will be codified in the near future, in order to represent all the grammatical categories and a small group of proper names will also be included to enrich the resource. Particular attention is paid to the Latin words that are used in everyday spoken and written Italian, for instance virus, medium, video, etc. The management tool of the database allows to visualize each Latin word inserted in the database, the Italian translation with its definition, the link with the English language, and slots showing additional information on request about its etymology, history, derivates, declension or conjugation. The Latin concept is represented as a node in the net of connections in its semantic field.

Se presenta una demostración del trabajo realizado para crear un sistema de traducción automática entre el euskera batua estándar y el vizcaíno estándar (un dialecto del euskera). La estandarización y la generación de datos lingüísticos de ambas variantes, y los progresos tecnológicos realizados en Apertium, han facilitado el manejo de datos de forma más eficiente para crear este traductor de código abierto. This is a demonstration of the work procedure to create a machine translation system between standard Basque and Biscayan (a Basque dialect). The standardization and creation of linguistic data of both variants, and the technological progress made on Apertium, have helped to manage data on a more efficient way to create this open-source software translator.

A lexical semantic database containing terms belonging to the specialized lexicon of the maritime navigation and maritime transport was built according to WordNet/EuroWordNet model. Our paper present a project planning the enrichment of the terminological database by means of a set of images. A short description is given about a) the structure of the terminological database and the domain conceptual modelling; b) the various features of the database management tool, and, among all, the possibility of visualizing, on demand, the image which is associated with the term being sought, contributing to clarify and refine the meaning of the term, increasing its information and communication effectiveness.

According to euroWordNet/ItalWordNet model, a terminological database has been created which contains terms belonging to the specialized lexicon of the technical-nautical and maritime transport domain. We want to give prominence to the frequency of idiomatic expressions and methaphors coming from this cultural environment used in everyday language. Our lexicographic research aims at: i) analyzing the relationship between the maritime domain as "source domain" and the concepts described in the "target domain"; ii) structuring the semantic code of idiomatic expressions in the terminological database as well as the link with the equivalent or closest expressions in English; iii) checking the use and the frequency of this type of idiomatic expressions and metaphors in a large corpus of Italian contemporary language.

This paper reports on the enhancing of a maritime terminological database by means of a set of terms belonging to meteorology. The main phases of this research are described and the initial results outlined: the model of the terminological database, following EWN/IWN; the criteria used to build corpora of specialized texts to be employed as the source for term selection and extraction; the use of a semantic database (IWN) as source for exporting synsets to be coded in the terminological resource and as reference (WN 3.0) for comparing and evaluating synsets. The set of semantic relations useful for codifying new terms belonging to the discipline of meteorology is described; in addition to the semantic relations provided by the IWN model, new relations are introduced which are more suitably tailored to specific needs either scientific or pragmatic.

The users' demand has determined the need to manage the growing new technical maritime terminology which includes very different domains such as the juridical or commercial ones. A terminological database was built by exploiting the computational tools of ItalWordNet (IWN) and its lexical-semantic model EuroWordNet. This paper concerns the development of database structure and data coding, relevance of the concepts of 'term' and 'domain', information potential of the terms, complexity of this domain and detailed ontology structuring recently undertaken and still in progress. Our domain structure is described defining a core set of terms representing the two main sub-domains specified in 'technical-nautical' and 'maritime transport' terminology. These terms are sufficiently general to be the root nodes of the core ontology we are developing. They are mostly domain-dependent, but the link with the Top Ontology of IWN remains, endorsing either general and 'foundation' information, or detailed description directly connected with the specific domain. This structure seems to be the most appropriate to characterize the main conceptual schemas that people of the technical-nautical or maritime transport "world" actually use, namely activity plans, navigation management, etc. Also a set of acronyms has been codified to represent their ever increasing use in maritime terminology. Through the semantic relations linking the synsets, every term 'inherits' the IWN Top Ontology definitions and becomes itself an integral part of the structure. While codifying a term in the maritime database, the reference is at the same time allowed to the Base Concepts of the terminological ontology embedding the term in the semantic network, showing that upper and core ontologies make it possible for the framework to integrate different views on the same domain in a meaningful way.

The paper reports on the creation of a terminological subset belonging to the maritime lexical domain; this set of terms is structured according to the design principles of the generic Italian lexical semantic database ItalWordNet, i.e. applying the same semantic relations model and using the possibility of linking the specialised terms to the corresponding closest concepts in WordNet. The main characteristics of the lexical semantic database are also described and, in particular: i) the construction of the terminological wordnet; ii) the approach adopted to connect the terminological database to the generic one; iii) some issues about the mapping between the Italian terms and the Princeton WordNet; iv) a first outline of a specific maritime domain ontology.

En esta tesis se define una arquitectura para una federación de recursos léxicos de naturaleza heterogénea. Abordamos el problema de cunsultar diversas fuentes léxicas heterogéneas desde el punto de vista de la integración de datos, utilizando un lenguaje de consulta unificado. In this work we define an architecture for a federation of highly heterogeneous lexical information sources. We address the problem of querying very different existing sources of lexical information from the point of view of information integration, using for this purpose a unique and common query language.

Al irse multiplicando las bases de datos léxicas en diferentes formatos, ha crecido también la preocupación acerca del nivel de reusabilidad de los recursos léxicos. El intercambio e integración de datos, así como el desarrollo de software común, son importantes para evitar duplicaciones de esfuerzos y posibilitar el desarrollo de bases de datos de información lingüística a gran escala. Es necesario un enfoque unificado en el desarrollo de herramientas para la creación y gestión de bases de datos léxicas. Su estructura debe seguir unas pautas basadas en unos niveles de abstracción claramente delimitados y su implementación debe estar controlada por métodos bien fundados de ingeniería de software. La tecnología de bases de datos relacionales es adecuada para soportar los datos y lograr una mayor flexibilidad, reusabilidad y expandibilidad. Con esa metodología se han desarrollado herramientas para la creación y gestión de diccionarios monolingües basados en una taxonomía conceptual con una única relación en forma de grafo orientado acíclico, donde una categoría puede tener más de un padre. Es éste un paso más en el camino que se está recorriendo hasta llegar a herramientas que gestionen bases de datos léxicas bien estructuradas y con múltiples relaciones semánticas. As lexical databases have proliferated in multiple formats, there has been a growing concern over the reusability of lexical resources. The interchange and integration of data, as well as the development of common software is increasingly important to avoid duplications of effort and enable the development of large-scale databases of linguistic information. A unified approach for the development of tools to create and manage lexical databases is necessary. Their structure must follow some patterns based on some clearly delimited abstraction levels and the implementation must be controlled by sound and well-founded software engineering methods. Relational database technology is suitable for data support and to attain a greater flexibility, reusability and expandability. With such a methodology we have developed tools for the creation and management of monolingual dictionaries based on a Directed Acyclic Graph shaped conceptual taxonomy with a single relation where a category can have more than one parent. This is a new step in the way to develop tools for managing well structured lexical databases and with multiple semantic relations.

This paper explores the automatic construction of a multilingual Lexical Knowledge Base from preexisting lexical resources. We present a new approach for linking already existing lexical/semantic hierarchies. The Relaxation labeling algorithm is used to select --among all the candidate connections proposed by a bilingual dictionary-- the right connection for each node in the taxonomy. We also propose several ways in which this technique could be applied to enrich and improve existing lexical databases.

One of the major current requirements in real world natural language processing applications is the construction of suitably structured and sufficiently exhaustive computational lexicons or lexical knowledge bases (LKBs). However, this is an extremely difficult, time consuming and expensive task, especially if it is necessary to start from scratch. For this reason, in recent years, researchers have begun to examine the potential or already existing resources, such as machine readable dictionaries and text archives, to see whether it is possible to develop procedures that can capture different kinds of lexical information from them, representing it in a computationally tractable and reusable formal framework.

This paper describes a method for the semi-automatic linking of lexical databases (LDBs) based on different source dictionaries and their partial merging. Equivalent entries from different dictionaries are mapped together and links are created between them at the sense level. The results can be modified interactively by the user and saved to form part of a new merged LDB whose entries will contain all the information included in the separate source LDBs. The aim is to provide a tool which makes it easier to compare and study lexical data derived from different sources and also to permit linguistic and lexical analyses of much richer data.

We describe the results of a syntactic-semantic parser for Cobuild dictionary definitions. Unlike previous work on the automatic analysis of machine readable dictionaries, the particular structure of the Cobuild definition allows us to derive information that classifies the lexical item mainly in terms of the selectional restrictions or preferences encoded on its arguments. The resulting formalized lexical entries contain data that has generally been lacking in other lexical representations but which is expected to be very useful in a wide range of NLP purposes. We show how this information can be used in dictionary sense disambiguation by creating links throughout the lexicon both on the paradigmatic and the syntagmatic axes.

The current state of progress of a research project for the design and development of a bilingual, Italian-English/English-Italian, lexical database system is presented. The aim is to create an integrated system in which a number of monolingual electronic dictionaries and/or lexical databases can be linked through the medium of a bilingual database. In addition, procedures are being implemented to establish access paths from the dictionary data to archives of texts in machine-readable form and language reference corpora, and vice versa. The system not only provides the standard look-up functions offered by conventional mono- and bilingual dictionaries but the organization of the data on database structures makes it possible to access and exploit the lexical information in many different and new ways. The structuring of the bilingual component is described in some detail and some of the possible applications envisaged for a tool of this type in the fields of pure and applied linguistics, lexicography and language learning are mentioned.

The digital era has brought some challenges to lexicographers, but it has also brought new opportunities as part of the rise of information technology and, more recently, the emergence of digital humanities. This paper provides a description of LeXmart, the framework that supports the digital development of the Portuguese Academy of Sciences Dictionary. LeXmart is a smart tool framework to support lexicographers' work that offers different types of tools, ranging from a structural editor to a set of validation tools. Given that the dictionary is stored in eXist-DB, LeXmart is developed on top of its ecosystem, using W3C standard languages, and offering default functionalities offered by eXist-DB, namely a RESTful API., H2020 - Horizon 2020 Framework Programme (731015)

This paper provides guidelines and examples for visualising lexical relations using Formal Concept Analysis. Relations in lexical databases often form trees, imperfect trees or poly-hierarchies which can be embedded into concept lattices. Many-to-many relations can be represented as concept lattices where the values from one domain are used as the formal objects and the values of the other domain as formal attributes. This paper further discusses algorithms for selecting meaningful subsets of lexical databases, the representation of complex relational structures in lexical databases and the use of lattices as basemaps for other lexical relations.