Introduction

Data interoperability is crucial for efficiently utilizing diverse datasets from multiple sources, enhancing data analytics, and driving insights across various fields. By ensuring that data is formatted and structured in ways that are universally understandable and accessible, we can leverage this data for more comprehensive analysis, decision-making, and innovation. Ontologies play a pivotal role in this context. They provide a structured framework that defines the relationships and categories within a particular domain, enabling data from different sources to be integrated and understood in a coherent manner. Ontologies ensure that data is not just interoperable at a superficial level but is semantically aligned, meaning that the data’s underlying meaning and context are consistent across different systems and datasets.

Less formal approaches like classifications and controlled vocabularies also contribute to interoperability. While they may not provide as deep a level of semantic alignment as ontologies, they offer a structured way to categorize and label data. When these classifications are mapped across different systems or datasets, they enable a basic level of interoperability that can significantly enhance data utility and integration. These mappings are the core topic of this section of the Foresight Portal.

Besides mapping unique elements of a classification to unique elements of different classification used in a different dataset (type 1 mappings), we can also create mappings between different levels of aggregation (type 2 mappings). The third type of mapping (type 3 mappings) defines relationships between unique elements of different concepts.

Geo-spatial mappings

Non-spatial data usually does have some sort of geographic or geo-spatial tag. If only at the level of identification of the country for which the data is relevant. GADM, the Database of Global Administrative Areas, is a high-resolution database of country administrative areas, with a goal of “all countries, at all levels, at any time period.” This forms a good starting point for making mappings to other classifications.

  •       •FAO statistics country (and territory codes), and region codes
  •       •World Bank statistics country (and territory codes), and region codes
  •       •ISO 3-letter country and territory codes
  •       •ISO 2-letter country and territory codes

Mapping food-items

There are different classifications related to food items. The food balance sheets of FAO use a certain classification of food items. This is different from the vary granular food item classification in nutrient content databases.

  •       •FAO Food Balance sheet food item classification
  •       •FAO production level commodity classifications
  •       •USDA nutrition table food items

Final remarks

Incorporating mapping information into dataset metadata is crucial. Metadata acts as a data guidebook, providing essential context, provenance information, and structural details. When metadata includes interoperability information like links to ontologies, classification details, or mapping information, it greatly enhances the dataset’s usability, accessibility, and value. This practice ensures that when data is shared or integrated, the necessary context and connections are available to utilize it effectively, fostering broader collaboration and more informed decision-making.