Historisierte Ontologien für „Linguistic Linked Open Data“-Ressourcen des Mittelalters

Abstract

Historical linguistic resources store valuable information about words and their meanings. These re­sources often suffer from limited online accessibility due to varied data formats, historical language stages, cultural contexts, etc. Modelling these resources as Linked Open Data (LOD) can overcome these challenges. Linguistics increasingly creates LOD using domain-specific ontologies for labelling linguistic elements such as words and multi-word expressions. For data access based not (only) on words but on their meanings, one must link these meanings to an extra-linguistic knowledge base describing the things of the world (lexical-semantic mapping, LSM). This enables a lexical-semantics-based data retrieval that is essential for asking cultural-historical questions of linguistic resources across various language stages. However, knowledge bases are typically ahistorical, reflecting the modern world; but historical concepts require historicised ontologies captur­ing the specificity of historical explanation patterns to avoid anachronistic mappings. Thus, the development of historicised ontologies, as demonstrated with ontologies for medieval medicine and law conducted by the research project ALMA, is crucial. Used for the LSM and LOD modelling of historical linguistic resources, these ensure the temporal and cultural context of the data is preserved, facilitating accurate and meaningful data queries.