LexIA
Open source tokeniser for the Spanish language with the capacity to contextualise legal texts, and to adapt to other sectors such as medicine, optimising the processing of compound terms.
LexIA is a project that seeks to develop a specialised tokeniser for the Spanish language, capable of identifying compound terms such as ‘Tribunal Supremo’ in a unique (one token instead of two) and efficient way.
It aims to improve the processing of legal data and to adapt to other industries, such as medicine, through customisable configurations. This tokeniser will be able to be integrated into natural language processing (NLP) workflows, optimising the speed and accuracy of legal text analysis.
The tool will be open source, facilitating its adoption and customisation by developers and companies.
- Complexity:7/10
- Technical specifications:Hugging Face Transformers, spaCy, NLTK, FastAPI, ANCORA
- CATEGORY 2024-2025 , Salamanca Univ.
- TAGS Android , Python , Pytorch





