The data model of an OES application specifies the types of content and metadata that exist, as well as their relationships with each other. Figure 2 shows a typical data model that, in our experience, is sufficient for the most reference works.

Central content types

• Articles (lemma in Fig. 2) form the core of every reference work. They contain text and multimedia elements (images, audio and video), as well as additional information such as titles, publication dates and DOIs. They are also linked to contributors and other content types.
• Contributors provide information about the people who contributed, such as their names, contact details and standard data. Contributors can have different roles, e.g. author, translator or other.
• Other typical content types include glossary entries to explain technical terms, bibliographic entries (literature) for sources and references, and register entries (index in Fig. 2), which facilitate targeted access to content.

All content types can be linked to each other.

Taxonomies and additional functions

• Taxonomies help to group and structure content. For example, they can be used to categorise articles (article types) as ‘reference work’ or ‘OES application’, and to tag articles with keywords for flexible tagging.
• Versioning and translations enable multilingual or continuously updated content to be maintained.