Through a series of posts over the last six months, I have been taking a run at some very basic questions about what it is we mean when we talk about content and, more particularly, about managing content. This might strike some as pedantry of the worst kind and perhaps it will be unmasked as such at some point in the future. The truth is that I simply don’t believe we, as a community of practitioners, have really dug deeply enough into the question of what is content for us to be able to move onto endeavouring to manage it effectively. And I think that my interest in this area has been fuelled by contact with a group of people who have begun to meet and discuss the topic of “intelligent content”. You can immediately see how the declaration that content can be intelligent would prompt me to first try to establish what content is and only then to proceed to consider the ways in which it might be considered “intelligent”.
I am returning to this topic in part because I am turning my thoughts to preparing my presentation for Intelligent Content 2010 which is being convened in Palm Springs on February 26-27, 2010. Last year, the inaugural meeting of this event was a great success in that it assembled a very engaging group of people, hailing from an unusually broad range of backgrounds, to collaboratively explore how content intelligence might be maximized and leveraged. In this year’s presentation, I will be elaborating on the idea of “Intelligent Content Management” as the ascendant and even necessary successor to “enterprise content management”, which I will suggest has somewhat under-performed both as an area of focus and as an industry.
For last year's event (www.intelligentcontent2009.com), I delivered a keynote address called "Content Fusion; or, There's a Piece of Data Lodged in my Document" (points were awarded to those who spotted the Monty Python reference).
In preparation for last year’s event, I also prepared a white paper that seemed to be reasonably well received by conference participants. A year later, I still find that it serves its intended purpose – that of introducing the concept of intelligent content and offering some thoughts on why it has arisen as a topic of interest at this point in time.
The following definition was put forward in this paper and it seemed to make it into circulation within the presentations and conversations of last year’s event, which I took to be a good sign that I was not too far off-base:
“Intelligence refers to the ability to acquire and apply knowledge (normally a quality attributed to people but not exclusively), or to a collection of information of value in a particular context (OED). Content can be considered intelligent when it expresses, in an open way, the full meaning underlying a communication such that the data, information and knowledge being expressed can be easily accessed and effectively leveraged by both people and the software applications that support them.”
Interestingly, the various explorations I have since pursued into what we mean by “content” have not undermined this definition in the least. In effect, I have been thinking more and more about the different ways in which data, information and knowledge manifest themselves in content and how each calls for different types of processing in order to be capitalized upon. For example, data is about representation and therefore it benefits from precision in its validation and interpretation. Information, oriented as it is towards transactional effectiveness, is geared to forming useful expressions of meaning and providing the mechanisms whereby these meanings can be rendered so as to achieve maximum effect. Knowledge transcends any one content instance although its traces will be evident throughout and in particular in the interrelationships, both explicit and implicit, manifest in the content that connect it to other content instances (forming a network, or web, of relationships that together represent an expression of a shared understanding of a subject). Intelligent content, as my original definition suggests, exposes, in a processable form, all of these content dimensions.
Needless to say, truly intelligent content is not an overly commonplace occurrence. This is not too surprising but nor should it be too distressing. In most environments, there will be a level of content intelligence that will recommend itself as sufficient and it will be the general capability of the community in which that content will be created and used that will determine what level of content intelligence is worthwhile pursuing. The challenge lies in determining where the right level of intelligence will be for a given community and then monitoring the environment for capability changes. Processes where the level of content intelligence is below what the community could leverage will under-perform, almost by definition. Processes where the level of content intelligence exceeds what the community as a whole can deal with will also under-perform in that the investment in content intelligence will not be balanced by the ability of process participants to leverage that intelligence.
So it is not true that elevated content intelligence will always be a good thing. However, with the general advancement of open content technologies (with this including the ongoing incursion of open standards into the technology mainstream) it is true to say that the time has arrived wherein elevated content intelligence can spark opportunities for explosive innovation amongst those in a position to exploit it.
See also:
In your definition, that phrase "in an open way" is interesting. Is it meant to indicate that the expression by the content of the full underlying meaning should transcend context?
(hmmm, that didn't come out quite right; I need an example... )
Consider a plain, untagged text file of "The Republic" sitting on a server, and then consider various readers who might open that file -- a philosopher, a classicist, a network technician (let's assume s/he's not a philosophy or classics enthusiast), a technical communicator, and perhaps others. Would you consider the content equally (un)intelligent across all these contexts?
Posted by: Milan Davidovic | January 20, 2010 at 08:57 AM
PS. I haven't read the full paper; perhaps if I went away and did that I'd find the answer there; it may be that "intelligence" of content has more to do with how non-human systems "perceive" the content.
Posted by: Milan Davidovic | January 20, 2010 at 09:01 AM
Hi Milan
This is a very interesting point you raise. I am reminded of the careful distinction, or qualification, that Claude Shannon made with respect to "information theory" - emphasizing that his use of the term "information" bore effectively no relation to the common usage of the term. As such, information theory is blind to the import or content of messages.
By a similar token, in discussing intelligent content we are looking at its technical accessibility and amenability to processing. This will ultimately have a bearing upon how people will perceive the content because it will determine the range and refinement of rendition options available. So whether it is Plato or Pulp Fiction, if it is in an unintelligent format (let's pick on MS Word), then there will be a limited number of ways I will be able to share it. This will materially impact how it is perceived, or at least how widely it is perceived. The question of the value or quality of the content's import does not, at least at this point, come into it.
If the value of the content is substantial, and this I would tend to align with what in my Content Fusion presentation I called the "semantic depth" of the content, then the value of elevated content intelligence will be higher (and with it the return on the investment needed to realize that elevated intelligence level). I would likely continue to ruminate on whether it would be possible to show that the sophistication of the patterns in high quality content will be higher (data, information and knowledge manifestations) and that this might actually provide some quantitative support to our preference for Plato over Pulp Fiction (with this preference being measured over time so that short term infatuation has a chance to dissipate). This might be an interesting path to explore or it could be a bog from which we may never escape. So in considering whether or not to continue, I will say "after you"...
Posted by: Joe Gollner | January 20, 2010 at 03:09 PM
Ah, I made it look as though I picked up on the wrong sense of "intelligence". Pulp Fiction, in fact, would have worked just as well (no shortage of semantic depth in Tarentino's work, I think; just of a different sort than Plato's). I was simply after an example from which different readers would extract different sorts and amounts of value (owing to their different background knowledge, experience, mental schema, and so on), and I may have been on my way to an analogy between those readers and the different technological systems that might encounter a piece of content.
But then I read your paper (good stuff, by the way) and I think I now have a better grip on "intelligence" as it relates to content.
Thanks...
Posted by: Milan Davidovic | January 20, 2010 at 07:37 PM
And I think I was being a little to glib in my response. I think that there are some very interesting paths to be followed by developing the analogy between the behaviour (perception and interpretation) of people and applications as they encounter content with differing levels of intelligence. In the act of encounter, there is definitely a sense in which the effective level of intelligence is being determined by the behaviour (or capability) of the interpreting "agent". On the other side of the equation, there is also a sense in which content exhibiting high levels of integral intelligence (if we allow ourselves to step over to that side of the equation) might be able to maintain decent levels of value across a wide array of agents (being able to work, as it were, with a range of capabilities). I think I have some more thinking to do about this. Thanks a lot for this Milan.
Posted by: Joe Gollner | January 21, 2010 at 06:28 AM