UI + Intrest Filtring +Talking to the Moon

this morning, as usual I begin by reading my RSS (on Google Reader ) and I I saw this Data Mining: Text Mining, Visualization and Social Media: The UI: The Killer App that No-one Will Build I really enjoyed this article.

I couldn't agree more with M. Hurst , here a quote from his article :

The obvious solution to this is to have a smart machine, right? This machine can cleverly tell you which of the atoms you should attend to. However, specifically when these communications are potentially of a personal nature (that is to say, from people that you identify, not nameless content spouters), it is hard to really trust the machine and accept that there is nothing in the dont-show-this pile that you need to read. Thus we fall back to the atoms and lumps model.

it was about email, the same apply to RSS Feed, or any information.

we are in a saturated information ERA, and fully and redundant information sphere .

we use information, and abuse it too.

to be in touch you have to subscribe to "too much information source".

the readability and usability of the information don't reside only in it content quality, but in it quantity too.

many are subscribed to more then 1000 RSS, many receive more then hundred email per day.

people in the search of being always in touch, always informed, and certain try to be the first informed saturated them selve by a huge amount of data that they can't use well.

it don't matter how many RSS (blog or new or any content) we are subscribed too , what really matter is how many we effectively read, how many of them take our attention, how many of them make us REACT.

passive information is a dead one.

information that don't make us REACT is the a loosely one.

the dream of a Intelligent machine that will sort the data to witch we are subscribed too, begin by a clearly definition of some entity, and one of them is:

"interesting information"

what make an information interesting or not, what make a human react to an information or not?

the answer to this question reside in a clear and precise definition of a personal interest.

"personal interest" is so complex, some try to define them as a suite of tag or label:

ex: "python", "django", "social media", "blogging"...

this is a suite of topic, a suite of topic of interest, people may define their interest by a suite of their topic of interest, but is it sufficient?

no.

not any "related information" to a topic of interest, is worth reading.

"personal interest", "topic of interest" aren't sufficient to determine a human interest.

an information may belong to a human interest, but if the source " the author" isn't in the "people of interest" then the information may be not worth spending time to read it or analyze it.

human interest may be defined using :

1)topic of interest.
2)people of interest.

the first define a set of topic which are interesting, where the second define a circle of trusted people worth reading.

and with a better User interface UI, that bring usability easiness of work maybe this would work.

but again to determine the "related information" intelligent algorithm have to emerge.

people trust other people, would people trust machine even the intelligent one?

until then let talk to the moon.

thank you M. Hurst .