Introduction to the Cadmus knowledgebase

This page provides an overview of the technologies in building the Cadmus knowledge base. Extensive tech documentation can be found in the upcoming white papers. The introductory page bites explanations of getting started when those technologies are.

Cadmus is named after the figure from Greek mythology.

What is history?

History is not just special people and special places and special times. It is many things causing many things to happen which cause many more things to happen.

The KB technology

A technical representation for this is graph databases, an established industry. Briefly, a graph database consists of nodes and the connections between them. For the circumstances for this Institute, the nodes represent specific entities, such as people or places or technologies, and the arcs represent the relationships between them. Not just the relationships, but also the reasons for the relationships.

Cadmus is the project name for the knowledgebase, and is being built on an RDF graph database. RDF is a specific type of graph database, and likewise is an estasblished industry.

The interface to Cadmus is through a conventional REST API, but the API is seriously not conventional, in a revolutionary way. This will be discussed more below.

Ingest

What kind of content goes into a history knowledgebase?

Clearly, people and places. And technologies. But the focus of Cadmus is on how ideas flowed. For example, not just what is an alpha channel and how does it work, but also who was involved in creating it? What concepts where they trying to establish in hardware? Where did those concepts come from?

Quite a bit of historical material already exists. Books. Interviews with people. Hardware and software products. Knowing that gives a start to the ingest process, i.e. bringing known information into a unified knowledgebase. For the most part this is well established technology.

However, one of the institute's mottos is "Using the technology of the last 50 years to tell the story of the next 50 years." So while it is a well understood process on how to intake currently practiced ways of expressing ideas, the future is coming, and there will be ways of expressing ideas that nobody has imagined yet.

At this point it's useful to bring in a couple of terms. Both are expressed  simplistically here, more detailed treatment is in the white papers:

Taxonomy – what do terms and phrases mean

Ontology – what is the definition of the relationships between entities

It's expected that Cadmus will have multiple taxonomies and ontologies, especially as research goes forward and it's better understood about how to define and implement them. As existing knowledge is ingested into Cadmus, it will be classified and organized according to the rules in the taxonomies and ontologies.

Venturi, the science in front of the knowledgebase


The number of nodes and arcs is overwhelming. It will easily be millions. At this complexity, it's not possible to draw a graph database on a screen. The only reasonable approach to visualization is XR.

What's conventionally called 3-D graphics is not 3-D - it is 3-D projected onto a 2-D surface. This process of projective geometry is lossy, i.e. information is lost in the process. Depth perception, for example. 3D graphics makes use of various depth cues. To suggest depth. However, humans exist in 3D. One might argue that humans perceive depth in 2D because of the flat image on the retina. However, humans start to develop depth perception from day one, but utilizing more than just vision, so it is not as limited as having only a screen.

In addition, a human's field of view is not constrained to a rectangle.

Given today's technology it's no longer necessary to be restricted to flat screens. This is where XR comes in. The various technologies present images that work very well at  making the humans human think that they are seeing an image with depth.

Two points:

- As mentioned before, human depth perception is not determined solely by vision
- XR tech is focused on technology first and having the humans adapt and map their human experience to the technology.

The Venturi science, named after the Italian physicist, reverses this approach. The human comes first; the technology must recognize what it's level of capability can do to leverage what the human wants, on the human’s terms.

The Vitruvian API

It's a lofty goal, but how to accomplish it. The key is that the API here, the Vitruvian API, Is an API to the human cognitive system. This is completely opposite to current practice where an API is to a technology.

Defining interactions in terms of how humans work means that humans don't have to go through any training, and most importantly, no mapping. Mapping is an opportunity for miscommunication. When technology comes first have to map what they are thinking to what the technology might be able to implement, and then map the results from that technology back to what the human was thinking. With the human first, the technology, which knows its own capabilities, then fits into the Vitruvian API. This puts the burden on the technology, but that's as it should be. The purpose of technology is to leverage what humans can do beyond what they can do as just humans.

Why the name Venturi?

A Venturi tube intakes air, modifies it, and ejects the modified air. The goal of the Cadmus knowledgebase is to ingest knowledge in pieces and fragments and provide technology and levers for the user to gain a greater understanding of the content.

Made in Elements