Forum > unintrusive semantic markup
I can largely agree. One thing that Twitter could do would be enable an additional out-of-band (OOB) payload for metadata. That would simplify the problem and allow the part of messages intended for humans to remain less cryptic.
The flip side of that, however, is that in-band (IB) data is flexible and fluid, and can be shared without any special client changes. It's also can be sent over existing communication channels, not just Twitter, like SMS and email. It's pretty interesting to be able to inject Twitter Data just by typing it in the message window--very useful for some use cases that we'll be talking more about soon.
My hope is that we'll see Twitter clients embrace Twitter Data and then transparently add IB data to messages, or transparently send additional messages with the IB data, thus making it OOB from the perspective of the user.
Todd
Todd Fast
"...transparently add IB data to messages, or transparently send additional messages with the IB data, thus making it OOB from the perspective of the user."
great point. all in all, i would rather have this embedded twitter data standard than not have it. and leaave it to the surrounding tech to provide multiple ways of handling the input/output.
a powerful notion is for crowd participation in specific campaigns where users are specifically told to use certain formatting/strings for tracking/parsing/interpreting... enhancing upon the simpler #hasttag.
sull
Todd,
thanks a lot for your twitterdata proposal!
I'm a Semantic Web researcher and I see a lot of interesting effects of your idea.
Yes, the "OOB payload" idea is very in line with what's going on in the semantic community, but I think that also the "IB data" can be very helpful: the idea of extracting structured data from unstructured text is definitely a hot research topic.
It's some time I'm looking for a way to understand "where do people tweet from?": if a twitter client embeds the $lat and $long data in the message, for example by using a GPS, extracting that piece of information by means of your twitterDataParse or similar stuff would be trivial (and it would provide me with the data I'm looking for...).
Keep up the good work and let me know if I can help somehow.
Irene (iricelino on twitter)
Irene Celino
Twitter is really a powerful tool for people nowadays.. wonderful posting..
S.JD
������������ � ������� �������� � �������. �� ������� ����� ����� ��������� ���������� �� �������� ������������� ������������� ������� � �������� ��������������, ������������ �������� ��������.
allergonaut
mail order brides white russia
<IMG>http://vfl.ru/i/20100902/dfbeaea0ad4ad7d12c4e3191a379d196_7.gif</IMG>
online dating agencies
sex russian mail order brides
russian girls hot
how to marry a russian woman
nasy russian girls
teenage russian girls nude korean chinese russian wives russian young women russian girls 12 to 16
Bactjagsjam
interesting, but would not semantic markup embedded within a tweet permalink be just as or more logical than structuring the actual message content? the UI (web, desktop/mobile client) could let a user apply metadata to a tweet... this data is used to markup the status permalink's html. the actual message remains natural language. services then just parse/scrape/interpret the semantics of a twitter status permalink. no?
would be more overhead, more resource intensive in regards to real-time.... but probably more natural to users while leaving messages looking normal and less cryptic. true, users are already using @ and # in messages so suggesting to users to also consider using $ to convey meaning to both humans and machines is not over the top. but i do think it still nudges a typical user to concern themselves with formatting rules as opposed to relying on UI forms during a post, which is what most people are comfortable with doing.
interesting stuff though. i'm not sure about it but i cannot argue too heavily against the effort either. can't hurt, that's for sure.