Part IV of FIVE EASY PIECES: Curation in a Federated System #content #Quora #DemandMedia #AI #semantics #ThinkState
So what exactly is curation in a federated system?
We’ve discussed why federation is important, but we haven’t yet discussed what it actually is.
Federation finds other instances and connects them, as opposed to instances working within themselves. In other words, it matches content, context, intent and action so that we don’t need to filter... We build and grow.
As explained by friend and colleague, Ishan Shapiro:
“Let's say we define federation as ‘aggregating without filtering’, in other words, the structuring of information without biases of algorithmic presentation of information.
Curation is storytelling using available knowledge and information. We create narratives through that information that can impart feelings, insights, results, etc. Curation has a conductive quality - curators act as lightning rods that ground the information, focus it and communicate it to other entities of the system. The qualities of good curation are relevance, salience, discernment and communication.
At the moment, curators are woefully under-equipped to communicate those stories or insights with the most relevance across communities due to the fractured quality of both proliferated information and of those of the core or extended communities themselves.
It can be said that this is all we can ever do - tell the best stories we can with the incomplete information that we have, and this is true. But what if we can better design ways to access that content and knowledge?
By using federated systems of information organization, we can vastly bootstrap our ability to curate relevant information, interact with it and communicate those insights through a wide spectrum of forms and mediums.
Further, content curation inside a federated system implies bringing together fractured communities. Communities at the moment are spread across platforms and tools, some with interoperability in the form of APIs, some not, with no guarantee that the content will stay on that platform or that the platform will continue to exist. Content is often on multiple platforms or services at the same time, fracturing the conversation and interaction with the content itself.
By dealing with the fundamentals of knowledge and content organization, we can provide the means to bring together these fractured communities and conversations, therefore allowing the communities themselves to better curate their own content and communicate that to a vastly larger audience.
We are all curators of our own content ecologies - by being given an infinitely stronger system to access and interact with the exaflood of content available to us and share our personal insights with others and be allowed access to their insights, we can vastly increase our ability to grow and learn, individually and collectively.”
The notion of content ecologies is quite powerful, but cannot be mistaken for content hubs or content farms. The natural tendency of media entities is to “syndicate” rather than federate content and relatable experiences.
One example of this is Demand Media’s model: it understands what search engines are doing, spreads out the costs for that content, sells more ads and then hedges the spread.
This is a smart business model given the current climate, but clearly it isn’t sustainable. In fact, it could be incredibly damaging to a market of journalists who now have little or no incentive to curate stories that are meaningful. And without good stories, there is no good content, no ads and therefore no revenue.
Demand Media is literally helping to cannibalize the very marketplace it chooses to play in.
Let’s use an example that seems to reside on the opposite end of the spectrum - Quora.
Quora is a wonderful proposition for the liberation of ideas and content, except for one, undeniable fact: it rests on a bedrock of cannibalized, unfederated media. Here, we can begin to see the ancillary challenges through a host of core questions:
- Does content moderation improve or intermediate (flatten) quality?
- How does sourcing of information scale, grow or self-manage?
- Where does content intend to go, and can it build onto itself?
- How are the associative experiences indexed into search?
- Who are the true experts?
- If there are none, what validates fact versus fiction?
- What does validity itself even look like?
- What is the difference between ‘accumulating’ knowledge and ‘discerning’ its meaning?
If semantics – what amount to pages of web content - are the catalytic elements, the “edges in between” (the Semantic Web), then we have some serious thinking to do around what word associations mean within the context of everyday life. Remember, the search paradigm is already hampered as it is. And severely hampered at that.
Does this mean that the semantics are not a stage-gate for the evolution of the Human Web?
Enter computational learning.
Empowering language & intent: the role of AI (artificial intuition) and computational learning systems.
Natural language challenges our ability to learn, adopt, adapt and process. This is a known fact.
The average word in the English language, for example, has roughly 16 different meanings when put into palatable context (ex: running a search query). This also means that the more we index, the more we search and the more confused we become... We literally lose meaning in the cycling of the very knowledge we seek.
The brunt of the technology and media systems we have today simply cannot correct this. When we look at curation, and its associative disciplines, this places us in a very precarious position.
As articulated by Chris Arkenberg in a Twitter conversation we had recently, what’s even more interesting is that SMS, hypertext and code or character-based systems (what used to be early language systems in the form of sigils, hieroglyphs, etc.) are fueling the compression of more info into less syntax.
We are literally creating a new meta-language. And the only way we can process it is by allowing machine learning to mimic behaviors, and allow machines to develop their own.
But, please, do not think of this as some doomsday Terminator or Matrix scenario.
Think of this as a means for providing strong, ethical, illogical objectivity to the way we are conditioned to think via more reductionist lenses. In other words, where we tend to get lost in emotion and spirit, or can’t articulate meaning through those elements, machines can help. They can help us transform.
This is where AI – artificial intuition - comes critically into play.
As stated by colleague, Monica Anderson:
“Computers currently do not understand semantics of human languages (e.g. in text). Attempts based on grammars, taxonomies, and other models of language or models of the world have failed because all models discard context, and in semantics, context is everything.
A new proprietary algorithm named ‘Artificial Intuition’ implements context-based understanding including perfect language disambiguation in conventional computers with large amounts of memory.
This completely new, fundamental capability will revolutionize document classification, data mining applications like threat detection and legal discovery, OCR correction, and speech understanding. Earliest applications support web search, spam filtering, and document classification.
This new algorithm also provides language generation, enabling applications like document summarization, dialog-based customer support, automatic copy editing and context-sensitive spelling correction, machine translation, and voice I/O for PIM and civil or private intelligence uses.
Extending language competence seamlessly into world competence leads to problem domain competent systems for computer-based document evaluation and curation, automated research assistants, and general knowledge federation systems.”
Hopefully, we can now see the reach and efficacy that federated systems can have – which is to change behavior and apply more universal frameworks for how we think, feel and act. It also means that stories are the linchpins for innovation.
In the last piece of this five part series, we will examine the role of transmedial thinking in the discovery of how storytelling can solve the challenges of Big Data and Big Business, and walk through what an interface experience might look like in the merger of content, intent and federated commerce.
The following illustrations manifested by Gavin Keech represent the infocology of how content develops as a fluid experience. Hopefully now these images carry more contextual resonance for you.
First, is the infocology of content as a fluid experience. While you can identify a pentagram shape within the design, do not be alarmed ;) Our intention is to build technologies around or representative of these flowcycles.
Here are the precepts, or experiential drivers, for content interface variables. Please note that these are explorations, frameworks if you will, that address content dynamics, but do not intend to identify all of them (as dynamics constantly change).
Clearly, this stuff is dense – we know this – but we want you to engage with us in a discourse around these elements. We live in a world of complex systems; embracing complexity is the bridge to intelligence discovery, and of course, an approach for curating experiences.