Entities
In Goetgevonden, the resolutions of the States General are offered as scans as well as transcripts. Various meaningful elements have been identified in the transcripts. We call these elements ‘entities’. The occurrences of the various entities have been extracted from the resolutions. The initial datasets that this yielded have then been curated.
At the time the REPUBLIC project was implemented, entity recognition in such a large amount of historical text as the resolutions of the States General was technically advanced. Because the extraction and curation of the entities (in view of the size of the material) are largely automated, the various entity datasets may contain errors. Occurrences of entities in the transcripts may have been missed or incorrectly linked to each other. It is good to take this into account when using entities to filter resolutions.
There are no set rules about what can be considered an entity. In Goetgevonden, entities have been chosen to help users search the resolutions. The following types of entities are distinguished:
References to other resolutions
All entities appear in the resolutions in multiple variants as a result of different spelling and writing methods. In addition, there may be errors in the automatic text recognition. This produces even more variants. In the curation process, each variant that appears in the text is linked to a standardised form of the entity.
In the curation process, most entity types are further subdivided into categories. The categories that are distinguished are always mentioned, when explaining the datasets with the individual entity types.
The entity datasets are also available as downloadable files.