Context | Mahesh CR

On Tagging & the Cognitive Processes

Marshall Kirkpatrick, in a very good post on how to fall in love with tagging, wonders why it has not taken off as expected. I have a thought on why this could be the case.

I have been using del.icio.us a little seriously of late and have faced a few challenges in tagging the content I come across. Before getting to details, I believe we all browse content on the web in a couple of modes i) Intent-driven and ii) Intent-less, defined as-

Intent-driven – Active when we have to accomplish something- say find a code-snippet, a quote, or a picture, and we seek out the appropriate site. We know what we want and go for it. Or even if we do not know where it is, we still know what it is we want (Perhaps this is why Google presents its bare search screen?!). When on the quest to find something everything else that is not ‘it‘ is a distraction.

Intent-less – Active when we are bored and just want to surf around. This is mostly driven by impulse and chance, even if an intent exists it is StumbleUpon exemplifies this mode, click a button and a random page is brought out for your perusal(Of course in StumbleUpon‘s case one has already expressed preference of topics to reduce the amount of noise).

Each of these modes, I believe, has a different execution process associated with it, though they use the same cognitive apparatus. Of course there is a continuum between them and its not an either-or situation.

Intent Ingredients
The quality of intent determines the extent of triggers that go off in our minds. Imagine, if you will, the neurons in your head lighting up in response to information that comes in. The more interesting and relevant the info the higher the number of neurons that glow. A handful of ingredients govern the number of flares that seem to go up in the head, and they are:

i) Attention – Or focus, this determines the extent of awareness we bring to the task. Keen observation is the outcome, not just to see things as they are but to better relate to the task on hand. Brings core and related concepts to the forefront by activating them, sort of setting the context. High when intent exists.

ii) Relevance to self – This is the ego element. That which does not have an impact on me fails to arouse my curiousity. However important and interesting, if its not relevant to me personally it carries no meaning. This influences the amount of attention too. This determines the emotional aspect of the equation. This acts as a catalyst to strengthen the associations between the concepts that light up.

iii) Immediate Utility – Sort of the ideal match to current task. Almost brings relief. This would perhaps peak the amount of neurons that fire up.

iv) Future Utility – This is the stuff we find along the way, not immediately relevant but we know it might be later. This could be based on situations we already encountered.

Each of the above factors influences how rich the process of cognizing is going to be, which directly determines the raw material from which we choose our tags.

And it is into this complex phenomena that we seek to insert the extra step of tagging. Reshmi Sinha covers the cognitive process of tagging in a very elegant way. Her observation that “tagging..taps into an existing cognitive process without adding add much cognitive cost”, is spot on.

But the cost is paid in another way, there is always a cost! The disruption in pausing the cognitive process to capture the tags and then resume the task undertaken is a chasm too wide for some of us to cross. And there in lies the challenge, the impediment to greater uptake of tagging systems.

Note: Most of the material on the web on tagging, including this stunning coverage on “Taxonomies and Tags” from O’Reilly’s Radar out here(select the February 2005 issue), consider tagging from an objective perspective; tagging used to describe content, classify and place it in a slot in a schema. However we need not consider tagging to merely capture what something is.

Tagging can also capture subjective perspectives. What do I feel about it? What will I do with it? It can be argued that in doing this tagging might no longer perform its original function and can diffuse the loose categorization scheme of tagging even further. But I feel this will encourage users to not bother overtly about where in the scheme of things the tag should fit and focus on what it means to them.

Disclaimer: Am not a scientist of anything! All comments are gathered from empirical observations, on a sample size of 1. 🙂

del.icio.us Tags: tagging, cognitive process, stumbleupon, delicious

Content Discoverability

There is a certain magic when Google comes up with what you were searching for. To a common user it has all the mystique that a poet possesses, to be able to peer within the mind’s eye and see how beauty and truth actually are without the encumbrances of form. This is perhaps the closest a machine can get in approaching the power of the human mind, if one takes into account just the ability to fetch information.

But one finds that Google and its ilk do not do “my” context well. Predominantly because what I am searching for, the motivations, the peculiar inflections of my need are known only to me. Any search engine can infer such context, as much as its made explicit in the information that I provide, only as it rises above the noise, by virtue of being manifested elsewhere by others like me. If am imperfect in my articulation of what I mean search engines falter. I have read that a bunch of search tweakers sit in front of their indexes adjusting the levers of the beast to account for context relevance. But this can only cater when my context is sufficiently replicated in others. Otherwise my need is just a nameless, faceless need that will not be catered for. Mind you this is no fault of the search engines.

That said it seems to me search engines have all the elegance of a brute force algorithm. And nothing more. Contextual relevance is increased by the loose taxonomies that community based tagging provides. It serves to carve out a specific form from the formless amorphous shapelessness that is information space of the internet. Where there was a chaos of ideas and opinions and facts, there is now a common mechanism where the user community can collectively label what a certain thing stands for.

All this works well for the internet, but what about my local content? What about the eBooks that I have? The mp3, photos, videos? As of now apart from the pathetic excuse that is Windows file system search the option is to use from the brute force approaches- Google for the desktop, Windows Desktop Live etc. As much as it gives me ability to peer deep into my content and get me what I want this does not make it easy for me to find things. When I search out to find something I don’t search for the sake of searching. There is a specific goal I want to achieve, towards which this search result is expected to help.

What I don’t have today is a taxonomy mechanism for my local content, a way of tagging with keywords that are relevant to me. Now note that these keywords might not have anything do with the content in the first place. What things mean to me is a function of my personality, my life and my environment.

A default taxonomy is available today which we are all extremely familiar with – the filesystem. With its file/folder nomenclature it allows us to classify content as we see it fit. But this model has serious flaws. First it confuses storage and context, is not flexible enough to keep the same content marked under multiple taxonomies and is static as in once defined its not possible to change it.

The next step for content search would be account for an individual’s context with greater fidelity than now. The crowd should influence but not entirely determine the outcome of my expectations from search engines and through it the internet.