I love diigo over all other similar services, but I do wish there was one added feature. I like del.icio.us' suggested tag feature, because no matter how much I wrack my brain for different tags, there's always a better one that someone has thought of. It'd be great to see what others have tagged on a certain webpage, sort of like a global consensus type thing on organizing bookmarks. Are there any features like this in the works?
There are several people working in this area: Patrick Schmitz has presented research into a model that works with both folksonomies and ontologies by leveraging statistical natural language processing. His goal is to develop a system that retains the flexibility of free tagging for annotation but make uses of ontology in the search and browse interface (Schmitz, 2006). Another proposal, from Dave Beckett (2006), is to make more use of the social context within which tags are created by separating the tool that creates the tags from the tool with which they are used. He also proposes that wiki pages should be created for individual tags which users could then add to/edit so that the wiki page, in effect, becomes the tag. The on-going process of refinement for each separate tag would form a kind of consensus as to the meaning of that tag and would also record the processes (the semantic path) by which the end result is being reached. This would, to take just one simple example, allow direct links to other language versions of the same tag.
In terms of bookmarking services such as Del.icio.us and the open source SiteBar (www.sitebar.org), one of the key problems is how best to classify the growing list of URLs. At the WWW2006 conference in Edinburgh, Dominic Benz et al, from the University of Freiburg, put forward an idea for automatically classifying bookmarks. The authors proposed an automated system which takes account of how the user has classified bookmarks in the past and how other people with similar interests have also classified their bookmarks. In other words find a similar user who has already classified and stored a bookmark and derive a recommendation based on what they did.
Seely Brown's book The Social Life of Information makes a powerful case for taking account and care of the social context in which information exists
Personally, I use the combination of Diigo Meta, Post to Diigo, Diigolet, my network and my library in a very careful order to avoid dataloss. Re: workflow, I'll post to a separate topic …
started by Joy Huang on 19 Apr 07
Keep of the great work!
There are several people working in this area: Patrick Schmitz has presented research into a
model that works with both folksonomies and ontologies by leveraging statistical natural
language processing. His goal is to develop a system that retains the flexibility of free tagging
for annotation but make uses of ontology in the search and browse interface (Schmitz, 2006).
Another proposal, from Dave Beckett (2006), is to make more use of the social context within which tags are created by separating the tool that creates the tags from the tool with which
they are used. He also proposes that wiki pages should be created for individual tags which
users could then add to/edit so that the wiki page, in effect, becomes the tag. The on-going
process of refinement for each separate tag would form a kind of consensus as to the meaning
of that tag and would also record the processes (the semantic path) by which the end result is
being reached. This would, to take just one simple example, allow direct links to other
language versions of the same tag.
In terms of bookmarking services such as Del.icio.us and the open source SiteBar
(www.sitebar.org), one of the key problems is how best to classify the growing list of URLs.
At the WWW2006 conference in Edinburgh, Dominic Benz et al, from the University of
Freiburg, put forward an idea for automatically classifying bookmarks. The authors proposed
an automated system which takes account of how the user has classified bookmarks in the
past and how other people with similar interests have also classified their bookmarks. In other
words find a similar user who has already classified and stored a bookmark and derive a
recommendation based on what they did.
Seely Brown's book The Social Life of Information makes a powerful case for taking account and
care of the social context in which information exists
See: http://www.informatik.uni-freiburg.de/cgnm/software/caribo/index_en.html
Noting with thanks that
* offers recommended tags
* in any capable browser
* without requiring installation.
Whilst Post to Diigo is good for gaining recommendations, it
overwrites existing tags.
Personally, I use the combination of
Diigo Meta, Post to Diigo, Diigolet, my network and my library in a very careful order to avoid dataloss. Re: workflow, I'll post to a separate topic …