Ward Cunningham, inventor of the wiki, is percolating another project: The Smallest Federated Wiki. This week he gave a presentation entitled, Missing From the Beginning: The Federation of Wikis Abstract, at the University of Advancing Technology (UAT) theatre, which is viewable here, and he’s been hosting Google+ hangouts about the work, too.

So, what is the Smallest Federated Wiki? The idea behind the work-in-progress, launched at IndieWebCamp this past summer and as explained here, is to innovate in three ways: The new Wiki shares through federation, composes by refactoring and wraps data with visualization. As Cunningham said in the March 7 presentation, “We’re making an ecosystem here for sharing data about ideas. I’m taking the conversation about how we’re going to live going forward, to be based on ideas backed up by data that we can understand because it has sensible visualizations.”

In the UAT Theatre Cunningham discussed this further, describing a vision of people flipping through web pages on their iPads and dragging and dropping content that’s meaningful to them to the wiki. And then, as they realize they’re on to something good – something that they want to share – they pull up those pages and further rearrange and edit and add to the information, within and across pages.

“Curating is the creative stuff of our future,” he said. “There’s just going to be so much stuff that curating it is the way we’re going to assemble the stories that are interesting.”

A Federated Wiki page will be a collection of paragraphs and paragraph-like items, such as images and data, constructed using a simple JSON model of the page. “What’s important is the basic and straightforward schema we have with this JSON can be shared between a wide variety of sites,” Cunningham has noted in one of his github video presentations exploring the idea. Even data collected from sensors can be published as a Federated Wiki page.

“This is really doing essentially Linked Data, just using JSON not RDF,” says Dr. Michael Hausenblas, Research Fellow at the Linked Data Research Centre at the Digital Enterprise Research Institute NUIG. “It doesn’t matter if you use JSON or an RDF serialisation, as long as you have fine-grained identifiers for each entity and you have typed links between entities. As long as you have this basic structure (of the Linked Data principles), you can one-to-one translate this to RDF.”  There may even be opportunity, he thinks, for the work to intersect with the Wikidata project, the new effort from Wikimedia that will help bring structured data into Wikipedia.

The Smallest Federated Wiki project also includes a journal feature, showing the series of edits that created the paragraph. An important part of the federated concept for the source page creators is that, “when you take my page and make changes, I can see what you’ve done and if I want to, [I can] change my site,” Cunningham says in another video. And if not, not. Edits done to a remote page become the curator’s, but servers can participate with each other so as to make it possible to get back to the original source as pages travel from site to site around the Federated Wiki World.

Also buried within the JSON code can be data, such as time series with temperatures. The capability to add visualizations to that data is there, too; for instance, with Stanford’s D3.JS library it is possible to reach into the documents on a page, find the data and render it into a chart. The visualization tools themselves can be pages within the Federated Wiki. “Those who visualize the data can be just another third party on this federated web,” Cunningham said in the presentation. Paragraphs also can use a variety of markup languages that come from plug-in support.

The interest and participation in the project so far has Cunningham feeling he’s on to something. During the presentation he told the audience that he thinks he’s “stumbled into a corner of the web space that is not well explored and where there’s a lot of opportunity.”