Versión en galego
Versión en español

Automatic and sustainable multilingual semantic tagger

ESMAS-ES⁺

ESMAS-ES+ is a project funded by Ministry of Science and Innovation/ State Research Agency/FEDER “A way to make Europe” (Project PID2022-137170OB-I00).

The project aims to develop and validate a combined method for designing a prototype for a semantic, multilingual, automatic and sustainable tagger for French, Galician, German and Spanish. Our proposal integrates lexical aspects ranging from computational lexicography and lexical semantics to technologies for the lexical analysis and for the formal representation of meaning, as well as their exploitation within computer applications.

Our semantic, multilingual, automatic and sustainable tagger requires a combined multimodular method, which should enable the automatic description and codification of the complexity of linguistic systems, especially the semantic information. Our starting hypothesis is that it is possible to develop this semantic tagger by applying a method that combines:

• a) combinatorial semantics, ontologies and lexical databases;
• b) techniques and strategies of Natural Language Processing (NLP; information retrieval and extraction, neural models and predictive methods),
• c) statistical techniques and
• d) tools designed for the automatic generation of language (NLG) and the automatic translation of lexical units.

This method therefore explores the development of a resource for semantic annotation and  automatic exploitation of large corpora in different languages —in other words, the same method and semantic techniques for different linguistic systems.

Key elements in our project are interoperability, integrability, sustainability, circularity of the linguistic and technological economy, as well as multilingualism.