Search

In this paper we describe the contributions made by the European H2020 project “Prêt-à-LLOD” (‘Ready-to-use Multilingual Linked Language Data for Knowledge Services across Sectors’) to the further development of the Linguistic Linked Open Data (LLOD) infrastructure. Prêt-à-LLOD aims to develop a new methodology for building data value chains applicable to a wide range of sectors and applications and based around language resources and language technologies that can be integrated by means of semantic technologies. We describe the methods implemented for increasing the number of language data sets in the LLOD. We also present the approach for ensuring interoperability and for porting LLOD data sets and services to other infrastructures, as well as the contribution of the projects to existing standards., This paper ist partially based upon work from COST Action CA18209 - NexusLinguarum "European Network for Web-centred Linguistic Data Science", supported by COST (European Cooperation in Science and Technology).

This article discusses a survey carried out within the NexusLinguarum COST Action which aimed to give an overview of existing guidelines (GLs) and best practices (BPs) in linguistic linked data. In particular it focused on four core tasks in the production/publication of linked data: generation, interlinking, publication, and validation. We discuss the importance of GLs and BPs for LLD before describing the survey and its results in full. Finally we offer a number of directions for future work in order to address the findings of the survey.

For Sign Languages (SLs), can we create a SignNet, like a WordNet for spoken languages: a network of semantic relations between constitutive elements of SLs? We first discuss approaches that link SL data to wordnets, or integrate such elements with some adaptations into the structure of WordNet. Then, we present requirements for a SignNet, which is built on SL data and then linked to WordNet.

This article provides a comprehensive and up-to-date survey of models and vocabularies for creating linguistic linked data (LLD) focusing on the latest developments in the area and both building upon and complementing previous works covering similar territory. The article begins with an overview of some recent trends which have had a significant impact on linked data models and vocabularies. Next, we give a general overview of existing vocabularies and models for different categories of LLD resource. After which we look at some of the latest developments in community standards and initiatives including descriptions of recent work on the OntoLex-Lemon model, a survey of recent initiatives in linguistic annotation and LLD, and a discussion of the LLD metadata vocabularies META-SHARE andlime. In the next part of the paper, we focus on the influence of projects on LLD models and vocabularies, starting with a general survey of relevant projects, before dedicating individual sections to a number of recent projects and their impact on LLD vocabularies and models. Finally, in the conclusion, we look ahead at some future challenges for LLD models and vocabularies. The appendix to the paper consists of a brief introduction to the OntoLex-Lemon model.

This paper describes an approach aiming at utilizing Wiktionary data for creating specialized lexical datasets which can be used for enriching other lexical (semantic) resources or for generating datasets that can be used for evaluating or improving NLP tasks, like Word Sense Disambiguation, Word-in-Context challenges, or Sense Linking across lexicons and dictionaries. We have focused on Wiktionary data about pronunciation information in English, and grammatical number and grammatical gender in German.

This paper presents Edie: ELEXIS Dictionary Evaluator. Edie is designed to create profiles for lexicographic resources accessible through the ELEXIS platform. These profiles can be used to evaluate and compare lexicographic resources, and in particular they can be used to identify potential data that could be linked.

We describe our current work for linking a new ontology for representing constitutive elements of Sign Languages with lexical data encoded within the OntoLex-Lemon framework. We first present very briefly the current state of the ontology, and show how transcriptions of signs can be represented in OntoLex-Lemon, in a minimalist manner, before addressing the challenges of linking the elements of the ontology to full lexical descriptions of the spoken languages

This volume documents the Proceedings of the 8th Workshop on Workshop on Linked Data in Linguistics, held on Friday teh 24th of June as part of the LREC 2022 conference (International Conference on Language Resources and Evaluation). Since its inception, the workshop series on Linked Data in Linguistics (LDL) established itself as the main venue for discussing how Linked Open Data (LOD) and semantic web technologies can be used for processing, analysing, publishing, and managing linguistic data. This includes the fields of natural language processing (NLP), language resources (LRs), lexicography and digital humanities (DH), and has been leading to the development of linguistic data science as a new area of study. The LDL workshop series has contributed greatly to the development of the Linguistic Linked Open Data (LLOD) cloud and the development of best practices for publishing and accessing language resources and providing language technology services on the web. Most notably, this includes community standards such as the NLP Interchange Format (NIF), the OntoLex-Lemon model of the W3C Community Group Ontology-Lexica, and numerous domain-specific adaptations and extensions that these models have had an influence on. In addition, there are an increasing number of national, European, and international research projects that build on LLOD technology. These will contribute to its further development and will help ensure the success of this workshop and a high attendance rate. The 10th anniversary edition of the LDL can count on the support of the COST action “NexusLinguarum: European Network for Web-centered Linguistic Data Science”, as well as two Horizon 2020 projects. Firstly, the Prêt-à-LLOD project, which is making linguistic linked open data ready-to-use, and, secondly, the ELEXIS project on building a lexicographic infrastructure. &nbsp

To facilitate corpus searches by classicists as well as to reduce data sparsity when training models, we focus on the automatic lemmatization of ancient Greek inscriptions, which have not received as much attention in this sense as literary text data has. We show that existing lemmatizers for ancient Greek, trained on literary data, are not performant on epigraphic data, due to major language differences between the two types of texts. We thus train the first inscription-specific lemmatizer achieving above 80% accuracy, and make both the models and the lemmatized data available to the community. We also provide a detailed error analysis highlighting peculiarities of inscriptions which again highlights the importance of a lemmatizer dedicated to inscriptions.

The scarcity of parallel data is a major limitation for Neural Machine Translation (NMT) systems, in particular for translation into morphologically rich languages (MRLs). An important way to overcome the lack of parallel data is to leverage target monolingual data, which is typically more abundant and easier to collect. We evaluate a number of techniques to achieve this, ranging from back-translation to random token masking, on the challenging task of translating English into four typologically diverse MRLs, under low-resource settings. Additionally, we introduce Inflection Pre-Training (or PT-Inflect), a novel pre-training objective whereby the NMT system is pre-trained on the task of re-inflecting lemmatized target sentences before being trained on standard source-to-target language translation. We conduct our evaluation on four typologically diverse target MRLs, and find that PT-Inflect surpasses NMT systems trained only on parallel data. While PT-Inflect is outperformed by back-translation overall, combining the two techniques leads to gains in some of the evaluated language pairs.

This article presents the first output of the Dutch FrameNet annotation tool, which facilitates both referential- and frame-annotations of language-independent corpora. On the referential level, the tool links in-text mentions to structured data, grounding the text in the real world. On the frame level, those same mentions are annotated with respect to their semantic sense. This way of annotating not only generates a rich linguistic dataset that is grounded in real-world event instances, but also guides the annotators in frame identification, resulting in high inter-annotator-agreement and consistent annotations across documents and at discourse level, exceeding traditional sentence level annotations of frame elements. Moreover, the annotation tool features a dynamic lexical lookup that increases the development of a cross-domain FrameNet lexicon.

Neural Network (NN) architectures are used more and more to model large amounts of data, such as text data available online. Transformer-based NN architectures have shown to be very useful for language modelling. Although many researchers study how such Language Models (LMs) work, not much attention has been paid to the privacy risks of training LMs on large amounts of data and publishing them online. This paper presents a new method for anonymizing a language model by presenting the way in which MedRoBERTa.nl, a Dutch language model for hospital notes, was anonymized. The two step method involves i) automatic anonymization of the training data and ii) semi-automatic anonymization of the LM's vocabulary. Adopting the fill-mask task where the model predicts what tokens are most probable to appear in a certain context, it was tested how often the model will predict a name in a context where a name should be. It was shown that it predicts a name-like token 0.2% of the time. Any name-like token that was predicted was never the name originally presented in the training data. By explaining how a LM trained on highly private real-world medical data can be safely published with open access, we hope that more language resources will be published openly and responsibly so the community can profit from them.

Topological Data Analysis (TDA) focuses on the inherent shape of (spatial) data. As such, it may provide useful methods to explore spatial representations of linguistic data (embeddings) which have become central in NLP. In this paper we aim to introduce TDA to researchers in language technology. We use TDA to represent document structure as so-called story trees. Story trees are hierarchical representations created from semantic vector representations of sentences via persistent homology. They can be used to identify and clearly visualize prominent components of a story line. We showcase their potential by using story trees to create extractive summaries for news stories.

We describe the status of work intending at including sign language lexical data within the OntoLex-Lemon framework. Our general goal is to provide for a multimodal extension to this framework, which was originally conceived for covering only the written and phonetic representation of lexical data. Our aim is to achieve in the longer term the same type of semantic interoperability between sign language lexical data as this is achieved for their spoken or written counterparts. We want also to achieve this goal across modalities: between sign language lexical data and spoken/written lexical data.

Can language models read biomedical texts and explain the biomedical mechanisms discussed? In this work we introduce a biomedical mechanism summarization task. Biomedical studies often investigate the mechanisms behind how one entity (e.g., a protein or a chemical) affects another in a biological context. The abstracts of these publications often include a focused set of sentences that present relevant supporting statements regarding such relationships, associated experimental evidence, and a concluding sentence that summarizes the mechanism underlying the relationship. We leverage this structure and create a summarization task, where the input is a collection of sentences and the main entities in an abstract, and the output includes the relationship and a sentence that summarizes the mechanism. Using a small amount of manually labeled mechanism sentences, we train a mechanism sentence classifier to filter a large biomedical abstract collection and create a summarization dataset with 22k instances. We also introduce conclusion sentence generation as a pretraining task with 611k instances. We benchmark the performance of large bio-domain language models. We find that while the pretraining task help improves performance, the best model produces acceptable mechanism outputs in only 32% of the instances, which shows the task presents significant challenges in biomedical language understanding and summarization.

LLOD (Linguistic Linked Open Data) is a generic name for a set of mutually connected language resources, using ontological relations. The connections between concepts and between concepts and their expressions in natural language make them suitable for both research and industrial applications in the area of content analysis, natural language understanding, (language- and knowledge-based) inferencing and other tasks. In the presented task, the concrete work will be on converting the SynSemClass project dataset (in part as a result of a previous Humane AI Net microproject called META-O-NLU) into LLOD, connecting it to the huge amount or interlinked data already available. A partner is involved in the Prêt-à-LLOD H2020 project, making this project synergistic in nature and multiplicative in terms of results in previous projects. Partners are also involved in the COST Action “European network for Web-centered linguistic data science” (NexusLinguarum).

This deliverable presents the second version of the strategic business development plan for Prêt-à-LLOD, developing and reporting the commercial partners’ specific business plans and achievements from version 1 of the report, delivered in month 18. The Prêt-à-LLOD commercial partners are: Semantic Web Company 1 (SWC) Oxford University Press 2 (OUP) Derilinx 3 (DLX) Semalytix 4 (SEM) All Prêt-à-LLOD commercial partners have been involved in the documentation of their respective business development plans and thereby in the creation of this deliverable with the aim of creating a comprehensive picture across the pilot projects of the Prêt- à-LLOD project. Updates to version 1 of the deliverable have been made in the following sections: 1.3.1 Impact on Technology companies 2.3 Business Model Canvas for pilot II – Linking Lexical Knowledge to Facilitate Rapid Integration and Wider Application of Lexicographic Resources for Technology Companies 2.4 BMC for pilot III - Supporting the Development of Public Services in Open Government both within and across borders 3.1 General market analysis 3.3.3 Competitor Overview: Pilot II - Linking Lexical Knowledge to Facilitate Rapid Integration and Wider Application of Lexicographic Resources for Technology Companies - Oxford University Press (OUP) 3.3.4 Market Analysis: Pilot II - Oxford University Press (OUP) 3.3.5 Competitor Overview: Pilot III - Derilinx 3.3.6 Market Analysis: Pilot III - Derilinx 4.1 SWOT Analysis for Prêt-à-LLOD 5 Collaboration with Academic Partners 6 Conclusion

This is the second deliverable reporting on activities pursued in the context of Task 5.4. “Community Building”, which is a part of WP5 “Language Resource and Service Sustainability”. While there might be some overlap with activities reported in the first version of this report (D.5.4), the focus is on activities taking place in the time between January 2020 and March 2021.

International audience; In this paper, we present the information extraction task proposed in the first round of the Covid-19 MLIA @ Eval Initiative. During this first round, we proposed to identify six categories of information potentially relevant for the Covid-19 issue (sign or symptom and disease, medical test, drug and treatment, legal rules, everyday life actions, and findings), on texts available in seven languages (English, French, German, Greek, Italian, Spanish, and Swedish). Since no gold standard annotations were given, the participants reused their existing tools and ressources. Four teams participated in this task.

OASIcs, Volume 93, LDK 2021, Complete Volume, OASIcs, Vol. 93, 3rd Conference on Language, Data and Knowledge (LDK 2021), pages 1-516

This paper describes ongoing work aiming at adding pronunciation information to lexical semantic resources, with a focus on open wordnets. Our goal is not only to add a new modality to those semantic networks, but also to mark heteronyms listed in them with the pronunciation information associated with their different meanings. This work could contribute in the longer term to the disambiguation of multi-modal resources, which are combining text and speech.

Front Matter, Table of Contents, Preface, Conference Organization, OASIcs, Vol. 93, 3rd Conference on Language, Data and Knowledge (LDK 2021), pages 0:i-0:xvi

This document provides a survey over vocabularies for language resources and services and sketch necessary extensions and the expected contribution of the Prêt-à-LLOD project to their further development for phenomena currently not sufficiently covered. Future updates with respect to this will be documented within Task 5.4. We focus on three main aspects of linguistically analyzed data 1. lexical-conceptual resources, i.e., repositories of terminology, lexical data, translation, and semantics, 2. linguistically annotated data, concerning linguistic analysis of textual or transcribed data, and 3. language resource terminology, i.e., linguistic data categories and metadata For these areas, we describe representative vocabularies from the Linguistic Linked Open Data community (RDF-based vocabularies) as well as other approaches (e.g., ISO TC37 standards), we identify a number of gaps, and we describe ongoing efforts to address these gaps within the Prêt-à-LLOD project.