In this paper we will present the research that is taking place at the NOVA CLUNL 1 where an international team is working on a financed project MORDigital 2 . MORDigital's goal is to encode the selected editions of Diccinario de Lingua Portugueza by António de Morais Silva (MOR), first published in 1789.
MORDigital's ultimate goals are, on the one hand, to promote accessibility to cultural heritage while fostering reusability and, on the other hand, to contribute towards a more significant presence of lexicographic digital content in Portuguese through open tools and standards. MOR represents a significant legacy, since it marks the beginning of Portuguese dictionaries, having served as a model for all subsequent lexicographic production. The team follows a new paradigm in lexicography, which results from the convergence between lexicography, terminology, computational linguistics, and ontologies as an integral part of digital humanities and linked (open) data. In the Portuguese context, this research fills a gap concerning searchable online retrodigitised dictionaries, built on current standards and methodologies which promote data sharing and harmonisation, namely TEI Lex-04 and Ontolex-Lemon5 . The team will further ensure the connection to other existing systems and lexical resources, particularly in the Portuguese-speaking world.
For this paper, after posing the theoretical background (terminology and lexicography) that /underpins our methodology, we will present 4 interrelated tasks:
1. Structuration of MOR's digitised versions using GROBID-Dictionaries6 , a specific software for the parsing, extraction and structuring of information extracted from dictionary text. In our case, the tool will be used to parse the constituent parts of each dictionary entry, which involves the preparation of a native encoding format that is compliant with the XML/TEI metamodel. 2. Presentation of a systematic analysis of the Mathematical Sciences and Medical Sciences domains, their related domain labels [6], [1] and other mechanisms, such as the use of formulae present in the definition which identifies the specialised field of knowledge. We will propose a hierarchical organisation that constitutes the foundation of domain ontologies. 3. Representation of the model in OWL resorting to Protégé7 , a free, open-source ontology editor. This means each class or individual in the ontology will be assigned a URI (Universal Resource Identifier), used to reference the label present in each of the lexicographic entries in accordance -whenever possible -with the TEI schemas. 4. Conversion of the TEI Lex-0 output of Task 4 into linked data using the RDF-based model Ontolex-Lemon; the conversion will be based on work already carried out in the scope of previous initiatives in rendering the two models more interoperable. The Ontolex-Lemon model has recently been extended by a lexicography module -lexicog8 -, which facilitates interoperability in modelling dictionaries as linked data.
At the end of the paper, we will discuss the results, highlighting the challenges that we faced.
This paper is supported by the MORDigital -Digitalização do Diccionario da Lingua Portugueza de António de Morais Silva [PTDC/LLT-LIN/6841/2020] project financed by the Portuguese National Funding through the FCT -Fundação para a Ciência e Tecnologia.