Caught in your History, marred in Garbage
Eli Pariser made a plea at his Junk Food Algorithms and the World They Feed Us TED lecture to Google, Facebook and other social networks not to cage the users (all of us) into their own history, and essentially make them unaware of different opinions.
This is similar to an analysis of Ghaddafi and his peculiar style of dressing and acting: Nobody does tell him anymore, that this or that might be questioned.
Whilst doing so might end up close to capital punishment, Google et al seems rather stuck in a wrong philosophical approach, that somebody can be defined by few mouse clicks.
I would argue even further. Google not only puts on blinds on my eyes, it is mainly garbage that I am being fed. Most of what I really want does not show up, because the algorithms operate without context.
For example, if I want to know something about "Formica" a group (genus) of ants living in the northern temperate region, I get Formica as material but furthermore a huge array of pages that somehow end up to have Formica in them ("About 6,770,000 results").
Does this make sense? Google is the victim of its own success, and I think a very stubbern company with the same symptoms of all the succesful companies: Becoming stupid because of the insistence of its own past, the search algorithms, web crawlers and server capacity to was at the begin of its success.
So, what is missing here is context in the search.
And this might not be just Google's mistake. As long as we stick to context insensitive html, we do not deliver Google the "Food for search" they need.
Alas, we need to move into the semantic tagging of content - and that's where I believe lies the success of our approach to add domain specific elements to publishing XML, such as the NLM Journal Archiving and Interchange Tag Suite, which then becomes TaxPub, which now is used by Pensoft with all its advantages of dissemination its content as widely as possibly, and being harvested. Such a small harvester is Plazi, that allows searching within the contest of treatments (the descriptions of species, such as species of Formica) or other elements that have been tagged, resulting for example in a list of species of a particular country
Eli Pariser made a plea at his Junk Food Algorithms and the World They Feed Us TED lecture to Google, Facebook and other social networks not to cage the users (all of us) into their own history, and essentially make them unaware of different opinions.
This is similar to an analysis of Ghaddafi and his peculiar style of dressing and acting: Nobody does tell him anymore, that this or that might be questioned.
Whilst doing so might end up close to capital punishment, Google et al seems rather stuck in a wrong philosophical approach, that somebody can be defined by few mouse clicks.
I would argue even further. Google not only puts on blinds on my eyes, it is mainly garbage that I am being fed. Most of what I really want does not show up, because the algorithms operate without context.
For example, if I want to know something about "Formica" a group (genus) of ants living in the northern temperate region, I get Formica as material but furthermore a huge array of pages that somehow end up to have Formica in them ("About 6,770,000 results").
Does this make sense? Google is the victim of its own success, and I think a very stubbern company with the same symptoms of all the succesful companies: Becoming stupid because of the insistence of its own past, the search algorithms, web crawlers and server capacity to was at the begin of its success.
So, what is missing here is context in the search.
And this might not be just Google's mistake. As long as we stick to context insensitive html, we do not deliver Google the "Food for search" they need.
Alas, we need to move into the semantic tagging of content - and that's where I believe lies the success of our approach to add domain specific elements to publishing XML, such as the NLM Journal Archiving and Interchange Tag Suite, which then becomes TaxPub, which now is used by Pensoft with all its advantages of dissemination its content as widely as possibly, and being harvested. Such a small harvester is Plazi, that allows searching within the contest of treatments (the descriptions of species, such as species of Formica) or other elements that have been tagged, resulting for example in a list of species of a particular country
posted by Donat Agosti at 1:30 AM
0 Comments:
Post a Comment
<< Home