@T: ALWAYS THE TWAIN

Well in this post I’m going to tackle the limitations of 5* star and other scalar rating systems, the Semantic Web Stack and sentiment engines, and provide a glimpse into how 360-2020 will effectively kill 3 big birds (that can’t take full flight) with a single stone. Plus 360-2020 will do what none of those 3 systems can do: take into account cultural, gender and other demographic factors.

All for fun on a quiet Sunday — LOL.

Here’s the Twain rationale for building 360-2020……………Enjoy…………

Whilst 5* star rating systems allow for the capture of 5 variables (typically — poor, bad, average, good, excellent), their lack of usefulness has very recently been admitted by no less than YouTube’s senior product manager:

There is also some suggestion for using the “thumbs-up-thumbs-down” or +1 / -1 or (percentage) marks out of 100.

Here’s the thing: NONE OF THESE METHODOLOGIES PROVIDE ANY CONTEXT ON WHY A USER IS GIVING THIS MARK OR HOW THEY ACTUALLY PERCEIVE IT.

Now, it’s also been written and said that the Semantic Web Stack should enable us to connect and contextualize content better because the data will be better structured (such as being able to differentiate Paris is a place, a person, a character from Homer’s Odysseus, etc.)

Here’s the thing: those classifications are still based on KEYWORD NOUNS (in case anyone hasn’t noticed) so whatever contextualization and link connecting of data objects remains restricted and rudimentary.

I’ve already noted the missing gaps in the Semantic Web Stack which I coined as a “Rubik cube of contextualization” (i.e., it’s rigid and not as flexible as if we adopted a naturalistic DNA approach) in my ‘The Global Brain’ Google knol:

* http://knol.google.com/k/twain-360-2020/the-global-brain-the-semantic-web-the/31fjy9fjsu1x2/19

Next up, sentiment engines which are said to be able to scrape data content not only for keyword nouns but also how a user’s text comments indicate they feel negative, neutral or positive about something.

Here’s the thing: sentiment engines are still producing results which are effectively -1, 0, +1 and whilst we can plot a table from the results there is NO way we can chart ongoing, real-time, dynamic N-dimension graphs wherein we can examine the constituents of those sentiments.

Below are examples from Twitter and Interaction London and the limitations of sentiment engines are covered in these blog links:

* http://threeminds.organic.com/2009/09/five_reasons_sentiment_analysi.html

* http://blog.techrigy.com/2009/10/five-myths-about-automatic-sentiment-analysis/

360-2020

360-2020 is a perceptions and values analytical system. It will have 3000+ variables to enable users to rate, review and rank content and it will produce graphs like this:

Yes, wherein possible code-wise, I am building it to simulate the way my brain perceives, contextualizes, cross-references, collaborates with, synergizes, deploys and evolves content. Hence my comment about a naturalistic DNA approach.

Yes, I am having to code parts of it personally because if I wait for so-called semantic tech leaders like that SemWeb CEO to “get” what contextualization reallys mean and should be able to do, then we will find ourselves in 2020 with Dr Larry Brilliant and Vint Cerf making the observation that “technology hasn’t changed that much since we built the WELL and TCP/IP”. That SemWeb CEO also lost my respect and confidence because he deleted my content, including my evidence that Google was already moving into the semantic space — analysis of mine that was about 12 months ahead of the tech media, actually.

WHAT’S IN A NAME?

It’s called 360-2020 for obvious reasons: 360 degree perspective and 20/20 acuity of vision. The temptation might have been to call it “TWAIN IT” the same way that Michael Bloomberg named his stock analytical system after himself, but the brand logo looks distinctive and says it all with the numbers. My friend GC loves it which is good.

Initially he noted that 5* star rating systems are so established and universally understood that people might not understand the need for 360-2020. Plus it’s not a name that’s readily familiar. I pointed out that Google, Knol, facebook, Wikipedia, Technorati and more are all invented names and before they became tech sector incumbents we didn’t even know we needed them either! Before Tim Berners-Lee originated the World Wide Web, someone invented the wheel and MS gave us something called a “browser” we didn’t realize we needed those or what the names were either!

Anyway, 360-2020 is informed by a lot of my direct experiences with technology and on the Web and I know it’s what’s needed for contextualization and consequence tracking.

IF CONTENT IS KING, CONTEXT IS QUEEN & CONSCIOUSNESS ARE THEIR PROGENY

To date, male coders and tech entrepreneurs have done a great job in encouraging the production and propagation of content (sharing, bookmarking, RSS, etc.). Context, though, will also need female input simply by virtue of academic research which notes that we understand communication nuances, emotions and relationship complexities differently from men:

* http://www.newsweek.com/id/203458

* http://www.medicalnewstoday.com/articles/168362.php

* http://ieet.org/index.php/IEET/more/treder20091022/

^{* http://roomfordebate.blogs.nytimes.com/2009/08/02/do-women-make-better-bosses/}

^{* http://www.medicaleducationonline.org/index2.php?option=com_content&do_pdf=1&id=46}

^{* http://www.ingentaconnect.com/content/bpsoc/joop/2007/00000080/00000004/art00010}

There’s no interest on my part to continue with any exacerbation of male-female differences or the statistical versus semantic arguments. My primary objective is to TWAIN what would initially seem to be two separate, silo and mutually exclusive areas and to develop the tools that can harness the best from each and both.

Now, THAT’s an interesting challenge — LOL.

Ok now I have to go out for Sunday dim sum………yummy!