Multilingual language models have been a crucial breakthrough as they considerably reduce the need of data for under-resourced languages. Nevertheless, the superiority of language-specific models has already been proven for languages having access to large amounts of data. In this work, we focus on Catalan with the aim to explore to what extent a medium-sized monolingual language model is competitive with state-of-the-art large multilingual models. For this, we: (1) build a clean, high-quality textual Catalan corpus (CaText), the largest to date (but only a fraction of the usual size of the previous work in monolingual language models), (2) train a Transformerbased language model for Catalan (BERTa), and (3) devise a thorough evaluation in a diversity of settings, comprising a complete array of downstream tasks, namely, Part of Speech Tagging, Named Entity Recognition and Classification, Text Classification, Question Answering, and Semantic Textual Similarity, with most of the corresponding datasets being created ex novo. The result is a new benchmark, the Catalan Language Understanding Benchmark (CLUB), which we publish as an open resource, together with the clean textual corpus, the language model, and the cleaning pipeline. Using state-of-the-art multilingual models and a monolingual model trained only on Wikipedia as baselines, we consistently observe the superiority of our model across tasks and settings.
We present research aiming to build tools for the normalization of User-Generated Content (UGC). We argue that processing this type of text requires the revisiting of the initial steps of Natural Language Processing (NLP), since UGC (micro-blog, blog, and, generally, Web 2.0 user generated texts) presents a number of non-standard communicative and linguistic characteristics -often closer to oral and colloquial language than to edited text. We present a corpus of UGC text in Spanish from three different sources: Twitter, consumer reviews and blogs, and describe its main characteristics. We motivate the need for UGC text normalization by analyzing the problems found when processing this type of text through a conventional language processing pipeline, particularly in the tasks of lemmatization and morphosyntactic tagging.
This paper presents an overview of a robust, broad-coverage, and application-independent natural language generation system. It demonstrates how the different language generation components function within a multilingual Machine Translation (MT) system, using the languages that we are currently working on (English, Spanish, Japanese, and Chinese). Section 1 provides a system description. Section 2 focuses on the generation components and their core set of rules. Section 3 describes an additional layer of generation rules included to address applicationspecific issues. Section 4 provides a brief description of the evaluation method and results for the MT system of which our generation components are a part.
METIS-II was an EU-FET MT project running from October 2004 to September 2007, which aimed at translating free text input without resorting to parallel corpora. The idea was to use "basic" linguistic tools and representations and to link them with patterns and statistics from the monolingual target-language corpus. The METIS-II project has four partners, translating from their "home" languages Greek, Dutch, German, and Spanish into English. The paper outlines the basic ideas of the project, their implementation, the resources used, and the results obtained. It also gives examples of how METIS-II has continued beyond its lifetime and the original scope of the project. On the basis of the results and experiences obtained, we believe that the approach is promising and offers the potential for development in various directions.
The European MAPA (Multilingual Anonymisation for Public Administrations) project aims at developing an open-source solution for automatic de-identification of medical and legal documents. We introduce here the context, partners and aims of the project, and report on preliminary results.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.