Search

This paper presents LexMeta, a metadata model for the description of lexical resources, such as dictionaries, word lists, glossaries, etc., to be used in language data catalogues mainly targeting the lexicographic and broader humanities communities but also users exploiting such resources in their research and applications. A comparative review of similar models is made in order to show the differences and commonalities with LexMeta. To enhance semantic interoperability and support the exchange of (meta)data across disciplinary and general catalogues, the most influential models for our purposes, namely FRBR (used in library catalogues) and META-SHARE (used for language resources), are selected as a base for the design of LexMeta. We discuss how these models are aligned and extended with new properties as required for the description of lexical resources. The formal representation of the model following the Linked Data paradigm aims to further enhance the semantic interoperability. The choice to implement it in two formats (as an RDF/OWL and as a Wikibase ontology) facilitates its adoption and hence its enrichment, yet poses challenges as to their synchronisation, which are addressed through automatic workflows. We conclude with ongoing and planned activities for the improvement of the model.

This article aims to provide an overview of current trends in vocabulary teaching, and to highlight several initiatives taken by the authors for teaching and/or improving French vocabulary for native learners (L1), and foreigners (French as second or foreign language, L2). After a cross disciplinary theoretical overview on the notions of ‘lexicon’ and ‘context’ we describe different activities having in common the necessity of considering the context in its broadest sense, that is, from a lexical, textual, discursive, and pragmatic perspective.1. Theoretical frameworkDifferent disciplines have been interested into the question of meaning, going from lexicology, semantics, to semiotics, and philosophy of language. Broadly speaking, the context is considered paramount, as Malinowski or Firth put it, for example through the famous quotation of the latter (arguing against structural linguistics) “you shall know a word by the company it keeps” (Firth, 1957). Taken to an extreme, Polguère (2015) calls it ‘lexical contextualism’, a conception that considers that words do not have meaning in the absolute (isolated): it would be their contexts that would bring it. Other less radical positions, such as the Generative Lexicon (Pustejovsky, 1995), suggest that all words have an original meaning. When polysemic, the different meanings derive by inference from an initial (original) meaning, depending on the usage of a word in a precise context. Contextualism, be it radical or moderated, is linked to the notion of ‘semantic potential’ proposed by Recanati (2004), notion that Polguère (ibid.) illustrates with the image of an unlit light bulb: “we should not argue against the fact that words do not ‘mean’ by themselves. They ‘have a meaning’, i.e. an inherent property that conditions their denotational effect in semiosis but considered by itself, a word is just like an unlit light bulb”. Last, but not least, Bakhtine considers the context in light of a dialogism inherent to the lexicon (kind of polyphony): the word “never refers to a single conscience, a single voice. (…) it is received through the voice of someone else” (Bakhtine, [1929] 2001, p. 235‑236).The notion of lexical meaning is at the heart of language learning and it can not be dissociated from the notion of context. As we can see, from contextualism to polyphony, the notion is complex and difficult to grasp. From there, how to facilitate access to meaning while teaching French L1 or L2?2. Pedagogical aspectsTo teach vocabulary (and to facilitate the access to word senses), language teachers can deploy explicit activities, that is specific and systematic exercises on vocabulary (Cellier, 2011). Implicit activities such as oral games, reading aloud, creating mind-maps, etc. to name a few, can also contribute to vocabulary learning. These implicit activities, which vary depending on the learners and on the level of language, have many advantages. A significant number of words are involved in effective classroom communication. In contextualizing them, the teacher can enhance their appropriation (Cellier, 2011). The role of the context is thus paramount. If quantitative approaches are important when learning vocabulary (increasing the lexical stock of the learner), qualitative approaches can not be ignored (Grossmann, 2011). While for a long time, pedagogical practices related to vocabulary have privileged quantitative aspects (word lists are still present in textbooks), a variety of activities are now implemented enabling to develop metalinguistic skills, i.e. getting access to word senses whilst working on contexts. Once again, the role of context is crucial.Recent studies in second language acquisition in general (Meara, 1996; Boulton, 1998) and specifically in French L2 (Cavalla, 2019) show the importance to teach/learn vocabulary in context as to dynamically build a diversified mental lexicon. The point is here to acquire skills and not lexical knowledge (Holec, 1994) to be able to cope with different situations, in comprehension and production.Besides, a variety of activities can be proposed in complement to textbook activities to handle heterogeneous classrooms: getting access to meaning may vary depending on the learner, his/her learning culture, his/her mother tongue (i.e. proximity with the languages already learnt), the languages learnt before, his/her motivation, etc. Insights on translation, cross-comprehension (Caddéo et Jamet, 2013), emotions (Cavalla, 2017), not to mention embodiment (Eshenauer et al., “Encorporer les langues vivantes” article in this issue), may also help memorizing new words and thus vocabulary appropriation. Finally, in communicative approaches for vocabulary teaching in L2, students are encouraged to use language as vehicle of communication and gradually develop their strategies in communication by using authentic texts. Vocabulary comprehension is linked to reflective activities on reading texts in the target language (Riquois, 2020) and on the role of context to infer meaning of unknown words.3. Contents of this articleIn this article, the authors propose different activities set up in French L1 or L2 classrooms, having in common a particular focus on the role of context in the processes of linguistic units’ learning. Following an overview of the domain as regards to its theoretical and educational aspects, we present innovative methods that go beyond more traditional approaches. Being complementary, those different activities are proposed as ‘food for thought’ on the complexity of contextualization for vocabulary appropriation and on the perspectives that contextualization offers on teaching/learning strategies. Considering the above, the authors raise the following question ‘is the process of accessing to meaning (word senses) really different in L1 and L2?’ and, consequently, they invite to rethink the artificial but persistent differentiation between L1 and L2 teaching approaches.

In the present article we make an effort to develop the ideas of the Russian linguist G.P. Melnikov about the determinants of the language system in the context of his study about the dynamic language nature. The article is devoted to argumentation of peculiarity of multi-structural languages lexical recourses based on their inner form. Significance of the present article is determined, at first, by increased interest from scientists to comparative study of the languages used within one territory in conditions of preservation of their national and cultural originality; second, by insufficient information about linguistic peculiarities of short folklore genre of the Teleut people; third, by scientific relevant tendency to explanatoriness of surveys in the sphere of modern linguistics that is directed to human beings and the world around them. The goal of the authors is to research typical communicatively relevant ways to reflect extralinguistic reality that determine vocabulary of the Russian and Teleut languages. To achieve the stated goal the authors use in the present article data-fetching methods (data from dictionaries, via informants interviewing and questionnaire), descriptions, comparisons, experiments as well as the methods of linguoculturological and statistic analyses. The material for this study is based on proverbs, superstitions, riddles, idioms that are regarded as the elements of the Russian and Teleut people spiritual culture. Communicative aspects of culturally significant facts comprehension identified through lexical universals analyses allow us to define the national world-image of the Russian and Teleut people. The conducted research proved that received during the analyses typical and unique coding methods of the information preserved in the depths of national memory determine lexical originality of compared languages. As a result, the study of paroemiological fund of the Russian and Teleut language that is based on the inner form category allowed the authors of the present article to reveal the mechanism of linguoculturological processes happening between them.

Distributional Semantic Models (DSMs) established themselves as a standard for the representation of word and sentence meaning. However, DSMs provide quantitative measurement of how strongly two linguistic expressions are related, without being able to automatically classify different semantic relations. Hence the notion of semantic similarity is underspecified in DSMs. We introduce Evalution-MAN in this paper as an effort to address this underspecification problem. Following the EVALution 1.0 dataset for English, we present a dataset for evaluating DSMs on the task of the identification of semantic relations in Mandarin Chinese. Moreover, we test different types of word vectors on the automatic learning of these semantic relations, and we evaluate them both in a unsupervised and in a supervised setting, finding that distributional models tend, in general, to assign higher similarity scores to synonyms and that deep learning classifiers are the best performing ones in the identification of semantic relations.

Lexical platform – the first step towards user-centred integration of lexical resources Lexical platform – the first step towards user-centred integration of lexical resources The paper describes the Lexical Platform - a means for lightweight integration of independent lexical resources. Lexical resources (LRs) are represented as web components that implement a minimal set of predefined programming interfaces. These provide functionality for querying and generate a simple, common presentation format. Therefore, a common data format is not needed and the identity of component LRs is preserved. Users can search, browse and navigate via resources on the basis of a limited set of anchor elements such as base form, word form and synset id. Platforma leksykalna – pierwszy krok w kierunku integracji zasobów leksykalnych zorientowanej na użytkowników Artykuł opisuje Platformę Leksykalną – sposób na lekką integrację niezależnych zasobów leksykalnych. Zasoby leksykalne są na niej reprezentowane jako komponenty webowe, które implementują minimalny zestaw predefiniowanych interfejsów programistycznych. Interfejsy te dostarczają funkcjonalność do przeszukiwania oraz generują prosty, jednolity format prezentacji zasobów. W związku z tym wspólny format danych nie jest konieczny i tożsamość składowych zasobów leksykalnych jest zachowana. Użytkownicy mogą przeszukiwać zasoby na podstawie ograniczonego zbioru odwołań takich jak forma podstawowa, forma wyrazowa i identyfikator synsetu.

This article investigates one kind of cross-part-of-speech relation for English, Czech and Zulu lexical resources in the form of semantic networks (wordnets). Many languages have rules whereby new words are derived regularly and productively from existing words via morphological processes. The morphologically unmarked base words and the derived words, which share a semantic core with the base words, can be interlinked and integrated into wordnets, where they typically form “derivational nests”, or subnets. Efforts are described to capture the morphological and semantic regularities of derivational processes in English, Czech and Zulu to compare the linguistic mechanisms and to exploit them for suitable computational processing and wordnet construction. While some work has been done for English and Czech already, wordnets for Zulu and other Bantu languages are still in their infancy. This article illustrates how Zulu can benefit from existing work.

This article contains the information about the defining and analysis of some rules which will permit the automatic retrieving of the romanian prefixed formations and enriching the lexical resources by automatic derivation with the romanian prefixes ne- and re-. There are also described some peculiarities of the romanian prefixes.

Word Sense Disambiguation Based on Large Scale Polish CLARIN Heterogeneous Lexical Resources Lexical resources can be applied in many different Natural Language Engineering tasks, but the most fundamental task is the recognition of word senses used in text contexts. The problem is difficult, not yet fully solved and different lexical resources provided varied support for it. Polish CLARIN lexical semantic resources are based on the plWordNet — a very large wordnet for Polish — as a central structure which is a basis for linking together several resources of different types. In this paper, several Word Sense Disambiguation (henceforth WSD) methods developed for Polish that utilise plWordNet are discussed. Textual sense descriptions in the traditional lexicon can be compared with text contexts using Lesk’s algorithm in order to find best matching senses. In the case of a wordnet, lexico-semantic relations provide the main description of word senses. Thus, first, we adapted and applied to Polish a WSD method based on the Page Rank. According to it, text words are mapped on their senses in the plWordNet graph and Page Rank algorithm is run to find senses with the highest scores. The method presents results lower but comparable to those reported for English. The error analysis showed that the main problems are: fine grained sense distinctions in plWordNet and limited number of connections between words of different parts of speech. In the second approach plWordNet expanded with the mapping onto the SUMO ontology concepts was used. Two scenarios for WSD were investigated: two step disambiguation and disambiguation based on combined networks of plWordNet and SUMO. In the former scenario, words are first assigned SUMO concepts and next plWordNet senses are disambiguated. In latter, plWordNet and SUMO are combined in one large network used next for the disambiguation of senses. The additional knowledge sources used in WSD improved the performance. The obtained results and potential further lines of developments were discussed.

Thesauri and similarly organised resources attract increasing interest of Natural Language Processing researchers. Thesauri age fast, so there is a constant need to update their vocabulary. Since a manual update cycle takes considerable time, automated methods are required. This work presents a tuneable method of measuring semantic relatedness, trained on Roget’s Thesaurus, which generates lists of terms related to words not yet in the Thesaurus. Using these lists of terms, we experiment with three methods of adding words to the Thesaurus. We add, with high confidence, over 5500 and 9600 new words and word senses to versions of Roget’s Thesaurus from 1911 and 1987 respectively. We evaluate our work both manually, and by applying the updated thesauri in three NLP tasks: selection of the best synonym from a set of candidates, pseudo-word-sense disambiguation, and SAT-style analogy problems. We find that the newly added words are of high quality. The additions significantly improve the performance of Roget’s-based methods in these NLP tasks. It compares favourably to the performance of WordNet-based methods. Our methods are general enough to work with different versions of Roget’s Thesaurus.

As underlined in Wierzbicka (1996: 239), a «linguistic definition is a scientific hypothesis about the concept encoded in a given word». The present paper focuses on the problem of what such hypothesis is about. Lexicon of any language is structured with a set of connections relating lexical items. That is to say that speakers of any language need to master not only separate words, but rather a cluster of connected senses. If so, then a lexical definition should lay out the concept the word lexicalizes in the way that gives access to the whole grasp of its conceptual, paradigmatic and syntagmatic relations. We discuss the topic of definitional structures developed in the context of the French Lexical Network (Réseau Lexical du Français).