Comentarios (4)

1 maxgano ha hecho un comentario el Enlace permanente

If I read you correctly, exploiting big data to its full potential will require a paradigm shift in how we manage and govern those assets. And, of course, it's really about managing and governing the practices and processes that utilize big data.

I was inspired by your opening comment about how we already use big data everyday, so please excuse the long comment that follows.
Yes, with a few keystrokes and mouse clicks, we can easily search assets around the globe. Of course, we are also traversing a world of data with a growing number of controls already imposed, whether it's Google's search algorithm, an organizations decisions on what to make public, or a government's decision on what to keep private or filter from a population of users.
All of this filtering and constraining of access occurs with little or no transparency to us, as users, seeking the information we need right now. And the combined impact of these controls creates an insanely complex warp to the information that appears on our screen. Have we seen everything? Can we trust the bias of whichever agencies have introduced filtering? At what point does that combined filtering render the exercise pointless?
And yet, with all it's imperfections, the system works and works brilliantly. I believe the reason for this is simple. People adapt, and they adapt quickly. They learn to read between the lines despite whatever algorithms and rules have been imposed in much the same way that friends living in Cold War Soviet Union once described their daily reading of Pravda. And like water, people will find a way around the controls as we have seen in very recent events of the Arab Spring.
What can we learn from this dynamic tension? In much the same way that the dialectic of Folksonomy versus Ontology is shaping how we organize information, the dialectic of Inquiry versus Control will define the discussion of how we govern information. There, I said it, now I just need to figure out what it all means !!!
Thanks again for a great post.

2 DataGovernor ha hecho un comentario el Enlace permanente

@maxgano, thanks so much for your insightful comments. I remember meeting a group of South African civil servants at an event in Joburg two years ago. The were discussing a draft bill to protect government information, which at the time was largely seen as a measure to restrict publication of information to the press that would reveal levels of ANC corruption.

Many journalists were apoplectic about the bill and I asked the group what they thought. They told me they weren't concerned about it, and I was shocked by that. They said it was just like the press controls during the Apartheid era, when the white minority government controlled information about the black majority and censored the press. "Information is like water in a river," they said. "Put some stones in the river, and the water flows around it. Dam the river, and eventually the Dam will break." They said they will find ways around the ANC law just like they did the Apartheid law.
I also like the "dialectic between Folksonomy vs Ontology." More comments please. You think and write very well.

3 paulerb ha hecho un comentario el Enlace permanente

Steven's response to maxgano suggests that we need to refine further the basic theory of information governance. Governance is the making and enforcing of rules for behavior within a domain whose members' interests are served by that system. It seems that the paradigm shift is in the "who," in the fluidity with which membership gets determined, and in turn whether the "behaviors"--of the data. of the users, of the emergent tools--serve the interests of an equally fluid system.

Two examples:
1. When Steven writes "Joburg," he creates a closed community of readers. The semantic coding of that term, the club its users belong to, sets limits on the coherence and availability of his information. Because I can comment on it here, though, I can open up new possibilities. So... the governance that allows rules to evolve continuously will be the most mature, and the most profitable for the members of that community who have decided what they value and wish that governance to protect and encourage.
2. The persistence of the water metaphor. The metaphor is meant to be limited, to point out that 1) water, like information, has immutable properties, and 2) you can't govern it without respecting those properties. But if we do not govern the metaphor, we discover new value. For example, the water dammed by rocks (hard rules, ontologies) may indeed flow around them, changing the banks of the stream. And from this evolving flow comes stream meander, and the creation of fertile valleys. We need not think of fixed rules as leading to revolutions. They may also lead to very intense and profitable behaviors...for everyone who engages in governance and can adjust to its evolution.

4 ErikSteinholtz ha hecho un comentario el Enlace permanente

...indeed, the question of governance of big data touches on the very profound question of how we make sense of the vast amounts of data we are exposed to.

I, in turn got inspired by the maxgano's comment on Ontology vs Folksonomy, as this is obviously at the heart of the "make sense" question. The fundamental issue in this contrast is that of central or distributed control. I think we can learn from economics, where a synthesis has developed over a couple of hundred years:
Adam Smith's invisible hand, illustrating the totally distributed control was the dominating theory until about a century ago, when Ronald Coase wrote the seminal "The nature of the firm", effectively providing a synthesis between centralised and distributed control as the way things actually work.
So... when semantics become increasingly challenging with big data, the solution for the interpretation may be a synthesis between Ontologies and Folksonomies, an "economy of information models" where the information model with the best ratio of content/integration cost (drawing on the Coase term transaction cost, if you care to read his work) will prevail. This may prove to be the "water" of the information models.
I realise I have strayed a bit from the topic.... but I hope I can be excused by the fact that semantics are integral to the value of Big Data.

Añadir un comentario Añadir un comentario