Unfortunately, you have to compensate most people (monetarily or notoriety etc.) to motivate them to annotate things correctly. As such, many content-centric sites have gone ahead and invented...
Unfortunately, you have to compensate most people (monetarily or notoriety etc.) to motivate them to annotate things correctly. As such, many content-centric sites have gone ahead and invented their own semantic annotation systems (which is something I think this article glosses over). It’s interesting to me that Twitter hashtags, Reddit/Tildes votes (and to some extent Tildes’ tags), Medium claps, etc. are essentially crowdsourcing semantic annotations (albeit in restricted domains—hashtags I suppose being less restricted) and they seem to actually be somewhat successful (I still don’t quite understand how people find hashtags fun…). The problem with crowdsourced annotations is that they are inherently unreliable. Crowd-sourced metadata is in need of better ways of building consensus that can’t be co-opted by moderators or other parties who would squelch legitimate annotations or corrupt the system by creating self-serving, but inaccurate annotations.
The article seems to jump around a bit between the semantic web being dead and giving what I think are good examples of how the semantic web has materialized and is growing. It's just happening a...
The article seems to jump around a bit between the semantic web being dead and giving what I think are good examples of how the semantic web has materialized and is growing. It's just happening a hell of a lot slower than expected and quite a bit different than the original ideas because the original idea was vague and full of problems. Assuming things continue how they are I think you'll see a lot of the functions the semantic web promised come into reality. Even their idea of providing access to your personal metadata can be accomplished today by signing into apps using Facebook.
I believe that's the key. There is no common standard. Or at least, no wildly adopted one. Who would start teaching people? Not 5 nerds at a conference, people who don't give a shit about the web....
Even if users were universally diligent and well-intentioned, in order for the metadata to be robust and reliable, users would all have to agree on a single representation for each important concept.
I believe that's the key. There is no common standard. Or at least, no wildly adopted one. Who would start teaching people? Not 5 nerds at a conference, people who don't give a shit about the web. Who would spear-head the annoying task of reminding people to adopt it, to correct mistakes and to build something useful enough to convince people it's worth the work? Before there even is certainty that it would work.
Google and facebook certainly are companies willing to take this on, but they want to keep the data locked into their systems. Maybe someone could come up with something that doesn't rely on ads. Please, someone come up with something that doesn't rely on ads!
Unfortunately, you have to compensate most people (monetarily or notoriety etc.) to motivate them to annotate things correctly. As such, many content-centric sites have gone ahead and invented their own semantic annotation systems (which is something I think this article glosses over). It’s interesting to me that Twitter hashtags, Reddit/Tildes votes (and to some extent Tildes’ tags), Medium claps, etc. are essentially crowdsourcing semantic annotations (albeit in restricted domains—hashtags I suppose being less restricted) and they seem to actually be somewhat successful (I still don’t quite understand how people find hashtags fun…). The problem with crowdsourced annotations is that they are inherently unreliable. Crowd-sourced metadata is in need of better ways of building consensus that can’t be co-opted by moderators or other parties who would squelch legitimate annotations or corrupt the system by creating self-serving, but inaccurate annotations.
The article seems to jump around a bit between the semantic web being dead and giving what I think are good examples of how the semantic web has materialized and is growing. It's just happening a hell of a lot slower than expected and quite a bit different than the original ideas because the original idea was vague and full of problems. Assuming things continue how they are I think you'll see a lot of the functions the semantic web promised come into reality. Even their idea of providing access to your personal metadata can be accomplished today by signing into apps using Facebook.
I believe that's the key. There is no common standard. Or at least, no wildly adopted one. Who would start teaching people? Not 5 nerds at a conference, people who don't give a shit about the web. Who would spear-head the annoying task of reminding people to adopt it, to correct mistakes and to build something useful enough to convince people it's worth the work? Before there even is certainty that it would work.
Google and facebook certainly are companies willing to take this on, but they want to keep the data locked into their systems. Maybe someone could come up with something that doesn't rely on ads. Please, someone come up with something that doesn't rely on ads!