Open Information Architecture
What about creating an “Open Information Architecture” plattform for everybody with “easy to use” tools? Open projects are the future of development not only inf information & communication technology. We should start with the base of knowledge: “Information” as soon as possible. Just an idea!
Ali Yeni | icsonx.com
I whole hearted agree. I have been working on something similar for the last 12 months. (+/- 10 years).
It’s a semantic contextual & computation engine. It’s RDF/SPARQL at core. It’s fully self contained. Downloads and self-installs, on a local machine or public server.
It is fully scriptable in VB, JS, AppleScript, Groovy, PHP, Python, Ruby, etc.
It supports Enterprise Integation Patterns. And a little BPML too. ie it is SOA-buzzword compliant.
It can curate documents including Word, Excel, PDF, JPGs. It retrieves meta from your email headers. And does some basic NLP on the content. It scans attachments too.
It can talk SNMP to network devices, and sniff meta-data (no data) from WiFi / Ethernet packets. It stores everything in RDF.
It has a rich web UI with nice widgets, validation. The UI is fully described by a custom ontology. It’s fully asychronous.
It has ASQ (Abstract Semantic Queries) that are easy to compose. They can CRUD SQL and SPARQL sources.
Using ASQ->ASQ transformations, it’s possible to read a SQL schema, transform each table into a grid, each field into a column. You can infer a rich UI from the meta-dta.
Of course, since it’s all RDF based everything is linked data, so it’s easy to ask: what data source does this widget visualise? That gives you “as-built” documentation.
So, we have an RDF-described MVC. It uses no public ontologies other than RDF and RDFS. It supports all of them. And can transform / infer between them.
In fact, I describe the “app” in triples, and it is executable and managed by the framework. Rich enterprise (and private) apps are just RDF. Easy to share.
It’s got graph analytics, to find shortest paths, centrality, clustering, etc. More work is need on matrix and graph analysis.
It’s written in Java, so its cross platform. It’s webstart so it installs in a few clicks. It’s a private knowledge server, that can scale to the enterprise, and beyond.
We’re just working on a few UI tweaks. Boring stuff, like a CRUD UI and a simpler ASQ query designer. It’s got a HTML and script editor too, so it self hosts it’s own web-based IDE.
Sounds amazing. Is there a demo available yet?
Sounds interesting.