The Horizon

When I first become obsessed with the internet I was very hesitant to try social bookmarking.  At the time, the only one that I had access to was del.icio.us.  I would sometimes end up on their website and get pretty flustered.  This initial learning curve is what turned me on at first.  All of that information stored in such a simple interface.  It was quite beautiful but I needed some learning before I could fully utilize the services.

The unique design scheme of social bookmarking is what attracts and deters people.  Social bookmarking has turned “tagging” into a near art form that has been gobbled up by anyone interested in creating a more powerful user interface for their web application or service.

Tagging allows for incredible customization.   Firstly, it does not constrain users as to what ways they can differentiate the content they are interested in.  For instance, let’s say that Jimmy wants to collect articles for learning the Python programming language.  Anytime he finds a website that serves this usage, he can add it to his social bookmarking webapp.  When he is asked to tag the website, he — the user — decides how he wants to organize his content.  Instead of a site giving him the popular ratings and usage of Python tutorials, he has the ability to set up experience levels (’beginner,’ ‘intermediate,’ and ‘advanced’) or perhaps tagging what particular aspect of Python is discussed in the article.  This allows social bookmarking sites to maintain near infinite versatility without the need for deciding what users will want to organize their content under.  Part of me wonders if the inventor of this versatile tagging interface was focusing on the DRY principle way too much and just coded a way for each tag to have a view of its own, but that’s besides the point: it’s a gorgeous use of technology and after I become acquainted with its ways, have become fully involved with social bookmarking, almost to the point of insanity.

Another useful feature that social bookmarking provides is that when Jimmy decides that he wants to add that Python article to his collection, most social bookmarking sites will give him a list of other popular tags that have been used to organize that same piece of content.   This allows for Jimmy to collaborate on the usage of specifc tags.  The nice thing about this is that not only does it give Jimmy ideas on how to organize his content, but not only that, the suggested tags are only suggestions!  To me, this is very important because it allows each user to decide what he/she thinks are good tags thus changing what the suggestions will be for the next user who wants to organize that piece of content in that user’s own collection.  Thus by the end of the day (metaphorically) you have a refined set of tags for organizing that piece of content.

Now after rattling off all of the things that I greatly enjoy about social bookmarking, I come to the real meat of my article: my qualms (should this sentence have more real content :P ).  This most versatile of tagging schemes provides nearly unlimited flexibility and turns the social bookmarking website into more of a “meta-website.”  The problem that I have with this laissez-fair attitude towards social bookmarking is that it doesn’t differentiate between tags.

For me, some tags should describe what type of content one is dealing with.  While other tags should describe other layers and facets of that particular content.  But the differentiation between “type tags” and the other tags (maybe even dividing other tags into genre, category, whatever you’re dealing with really) is null in most social bookmarking web services.  This to me is something important that I want to see added as a feature.  In a way I am asking for a way to turn some tags into pseudo-categories with subcategories.  I made a little chart on bubbl.us to illustrate an initital type-tag categorization system I am working on:

I know that del.icio.us has a way of grouping tags.  To be completely honest, before I started working on this article I had been using the grouping feature on del.icio.us to divide my tags by subject area.  I also had set up a ‘.inbox’ group with my ‘todo,’ ‘toread,’ ‘donereading’ tags in it.  I think I will keep this differentiation but I will also be adding a ‘type-tag’ group so that I can feel that my content has some heirarchical organization to it.  But I am not sure if this will be enough to satisfy me as it lacks any true differentiation other than a semi-superficial grouping.

Applying (or at least having available) a more strict taxonomy for types, genres, etc., will allow for users to understand content in better ways and find content in more ways.  Many people feel that the ultimate flexibility of a folksonomy is really the current apothoesis of organization but in reaility I think that it is but one feature of taxonomical organization.  It fluidly links a lot of content together and involves people immensely in the organizational process, but it limits semantic taxonomies and thus actually limits the amount of content it can be connected to.

For instance, let’s say you have a python tutorial.  In the folksonomy instance, Jimmy would add the site to his collection and then choose all the tags he wants to apply, having some based on what others deem popular and others based on what Jimmy wants to help him label his content.  In a situation where there is an additional tagging feature for types (or other semantic information), Jimmy would be able to add type-tags that say that the content is a resource, and more specificallly it is a tutorial resource.  Jimmy would then add content-tags that say that the content relates to python.

This differentiation of type-tags from content-tags could allow for more suggestions and more organization.  A web-service could suggest to a user some type-tags pulled from the content. I am sure that tutorials tend to contain words like ‘tutorial,’ ‘beginner,’ and many other words that obviously differentiate what type of resource it is.  Some other type-tags like opinion articles miight be more difficult but nothing that a user could not handle.

This additional heirarchical level of tagging would allow a more refined and easy search to be done as well.  After Jimmy adds the content he could be given suggested content that are specifically tutorials as well as providing other suggestion boxes should other kinds of type-tags (dividing up the content for the user).  Jimmy would see some more tutorials (possibly narrowed down to beginner tutorials as well).  The web service could also should intermediate tutorials in another block and cheat sheets and top10 lists of tutorials and other types of content all seperated so that the user can more easily understand what kind of content he will be accessing as well as finding right content that willhelp his research.  For Jimmy, advanced tutorials, cheat sheets, tips, and other more specific and advanced types of content won’t be useful to a beginner like Jimmy.

Trackback URI | Comments RSS