Structuring Knowledge

In this second model, the information and associations derived from the conceptual map are presented in a more abstract form. Elements such as individuals, dates, topics, and locations are treated as entities, and the links among them are shown as relationships.

A detailed assessment was carried out on the selected items, focusing on how each is described on its original source website. This involved closely analyzing the existing descriptions to identify the key attributes and features emphasized by the sources. Based on this evaluation, appropriate metadata standards were selected to best represent each object. The chosen standards align with the original descriptions, ensuring consistency and completeness in the metadata while accurately reflecting the way the items are presented by their respective sources.

In this section, the items were revisited to be characterized according to the attributes defined in the Conceptual Model. For each item, a table is presented with three columns representing the components of an RDF triple (subject, predicate, object), where each row corresponds to a distinct statement.

After identifying the relevant standards, this final stage enables a comprehensive description and interconnection of the selected items. The project’s overarching aim has been to ensure coherent and interoperable data delivery by adopting widely recognized semantic ontologies. To structure the information effectively, all properties were organized into four categories: WHO, WHERE, WHEN, and WHAT.

• Dublin Core: The Dublin Core Metadata Initiative (DCMI) Terms were selected for their broad applicability and established use in describing diverse types of resources through a standardized vocabulary.
• Schema.org: Schema.org was chosen for its extensive and widely implemented schemas, which support the structuring of data across the web and enhance interoperability.
• CIDOC CRM: The CIDOC Conceptual Reference Model (CRM) was adopted due to its suitability for cultural heritage contexts. As a domain-specific ontology, it supports the integration and contextualization of heterogeneous data from various institutions and disciplines—key requirements for this project.
• FOAF: The Friend of a Friend (FOAF) ontology was included to model information about people and their relationships. It is particularly useful for describing individuals, their roles, and their connections within broader cultural and historical networks.

In instances where no direct metadata alignment was identified, "N/A" was indicated in the table below to reflect the absence of a suitable mapping.

The theoretical model provides a comprehensive overview expressed in natural language, outlining the properties associated with each entity and the semantic relationships between them. These relationships—visualized through pink arrows—illustrate how individuals, places, objects, and events are interconnected within Gabriel García Márquez’s literary and biographical universe.

The model responds to fundamental questions that guide the structure of the Linked Open Data (LOD) network:

• WHO? Identifies individuals related to Gabriel García Márquez—both real and fictional—and specifies their roles (e.g., author, character, publisher, historical figure).
• WHAT? Describes the nature of objects such as manuscripts, books, characters, and concepts, detailing their characteristics, types, and functions.
• WHERE? Maps the locations connected to the author’s life, his works, and the archival documents, enabling spatial contextualization.
• WHEN? Registers key dates and time periods—such as birth and publication dates, historical events, and narrative timelines—supporting temporal navigation.

Through this structured approach, the LOD Gabriel García Márquez project enables users to traverse the author’s intellectual and narrative networks in a meaningful and machine-readable way.

Returning to the Entity/Relationship (E/R) model, additional layers of detail are incorporated to expand upon the structure previously outlined. This enhanced version aims to provide a more granular understanding of the relationships between entities and their associated attributes. By transitioning from descriptive natural language to a formalized schema, the model enables clearer identification of nodes (e.g., people, places, works, events) and the semantic connections that link them.

Ultimately, this formal mapping achieves the highest level of abstraction by adopting an ontological structure. It represents entities and their relationships using standardized classes and properties defined by authoritative institutions, while retaining literals when necessary. Clear relationships are established between subjects, predicates, and objects—whether entities or literals. The model is based on established schemas, ontologies, and controlled vocabularies, ensuring semantic consistency and interoperability with widely accepted Linked Open Data conventions.