Creation of a High-quality, Register-diversified Parallel (English-Spanish) Corpus for Linguistic and Computational Investigations

Abstract

This paper outlines current work on the construction of a high-quality, richly-annotated and register-diversified parallel corpus for the English-Spanish language pair, as currently carried out within the framework of the MULTINOT project. The corpus consists of original and translated texts in both directions and is designed as a multifunctional resource to be used in a number of disciplines such as corpus-based contrastive linguistic and translation studies, machine translation, computer-assisted translation, computer-assisted language learning and terminology extraction. The paper describes the structure of the corpus -which includes four subcorpora: English originals (EO) and Spanish originals (SO), English translations (Etrans) and Spanish translations (Strans)-, the registers selected for inclusion in the corpus, and the methodology used to guarantee the quality of the processing steps to enrich the corpus with linguistic information at different levels.

Stance and Engagement in English and Spanish Journalistic Texts: Towards a Reliable Annotation Scheme for Linguistic and Computational Purposes

“…In order to test the reliability and consistency of the core and the extended tagsets of the proposed annotation scheme, a pilot agreement study was performed by two expert annotators working on a small training corpus of eighteen comparable texts extracted from the larger set of bilingual newspaper texts contained in the MULTI-NOT corpus (Lavid et al 2015). The training corpus contained eighteen texts, evenly divided into comparable sets of English and Spanish news reports, editorials and letters to the editor.…”

“…It is necessary, therefore, to empirically test the validity of these categories not only for theoretical purposes but also to be used for the annotation of large datasets, which can be later used for computational purposes such as Machine Learning, Automatic Text Classification, Multilingual Information Retrieval and Sentiment Analysis, among others. In this paper, we undertake this task in the context of the MULTINOT project, aimed at the creation and empirical validation of discourse features in English and Spanish through corpus analysis and annotation (see Lavid et al 2015;Lavid, Moratón 2016).…”

“…The paper describes the process of validating a reliable annotation scheme for the categories of Stance and Engagement in English and Spanish using a bilingual sample of English-Spanish journalistic texts extracted from the MULTINOT corpus (Lavid et al 2015). The bilingual sample includes three different newspaper genres: news reports, editorials and letters to the editor.…”

A linguistically-motivated annotation model of modality in English and Spanish: Insights from MULTINOT

“…In this paper we describe the construction and empirical validation of a modality annotation scheme for English and Spanish in the context of the MULTINOT project, whose main aim is the development of a parallel English-Spanish corpus which is balanced -in terms of register diversity and translation directions -and whose design and enrichment with multiple layers of linguistic annotations focuses on quality rather than on quantity (see Lavid et al 2015) 2 .…”

Designing and Validating an Annotation Model of Dynamic Modality for English and Spanish: Issues and Problems

“…The MULTINOT corpus consists of originals and translated texts in both directions and is enriched with linguistic annotations which can be exploited in a number of linguistic, applied and computational contexts (see Lavid et al 2015).…”