RefInformationImport

As with any other database, one of HyperGraph's core purposes is to simply store data that can be retrieved efficiently. However, unlike conventional databases, HyperGraph is designed to manage complex information and it recognizes from the start that we are interested in recording data because of its informational import. That is, we are interested in the meaning and interpretation of the data. We record data because we need to manage information.

The distinction between data and information is intuitively well-understood. While the Webster dictionary loosely defines data as “factual information” and information as “knowledge obtained from” or “communication of such and such”, in science the two terms are better narrowed down. Data is usually understood to mean some raw, isolated, uninterpreted content in computer memory (or on paper), and information is interpreted data or, in other words, data placed in context and acted upon. Thus, the term information carries in its connotation the notion of semantics and semantics is only possible when there is some process involved, when there are some dynamics.

In HyperGraph, a more basic view is taken of the information concept. We say that atoms in HyperGraph represent information, or carry information in virtue of their relationships with other atoms, and in virtue of their types and values. This is more in line with the Latin etymology of the term which means “to put in form”. Atoms in HyperGraph are put in form by being typed and by being linked to other atoms. This is not to say that HyperGraph is really unique because it overlays a certain structure to some data. Standard databases also put raw data into form through mechanisms such as tables, primary key, foreign keys, constraints etc. However:

1. In standard databases there is only one meta-level, whereas in HyperGraph there's no limit to the number of structural levels because of the link-node duality and the potentially infinite type tower.
2. Also, in standard databases, the formalism for structuring data is fixed and immutable, while HyperGraph provides a much more general environment for applying different formalisms.
3. Most importantly, the fact that the informational import of a given datum is what really matters is explicitly recognized in HyperGraph and support is provided for working at the level of information.

The last point in the above list is what we wish to elaborate more on in this section.