D1.1 Mapping Report: Open Scientific and Specialised Language Resources in Europe

This report presents the data mapping activity conducted to identify scientific and specialised language resources available in 29 European languages. This task is part of the OSCAIL project, which aims to investigate the opportunities offered by Machine Translation (MT) and Artificial Intelligence (AI) to support multilingualism in scholarly communication. To assess this potential, the first step involves delivering fine-tuned models specifically designed to support the project’s three selected use cases: (i) peer review, (ii) discoverability, and (iii) accessibility of scholarly publications. As model fine-tuning requires specialised language datasets, the initial task of Work Package 1 (WP1) consisted in mapping the existing resources with a view to collecting, processing and re-using them when possible and relevant. By centralising these language resources, the project also seeks to contribute to the establishment of multilingual digital commons for scholarly communication. The digital commons represent a pillar for European digital sovereignty, enabling the creation of models and use cases anchored in open science, community-driven governance, and broad linguistic representation.