Hmm, I often use tags to navigate through posts on some very large sites, albeit, I don’t often find them any more useful than categories. Except on the sites you mentioned. The likes of flickr, delicious and wists are really made better by tagging.

Mark - I’m a category freak (probably was a librarian in a former life), but I don’t get much use out of tags. In flickr, they’re handy, but I wouldn’t miss the tags feature.

In del.icio.us I think it’s almost too easy to create new tags for yourself and you run the risk of going overboard. You could waste an awful lot of time micro-managing your link tags!

tags! It’s something we’ve also been wrestling with here at Glamorgan. So far we seem to be using them purely on the CMS side of things. For example, events tag cloud is an admin side view of our tagging of ‘events’. For the user side we tend to replace them with ‘expanded information’, e.g.  A page from our site, below the fold we provide a “you may also be interested in ..” section, which is tag driven.

For pure tag clouds we so far are only experimenting with it on our soon to be released blog portal, and there are certainly some reservations within the team.

I think the blog post (the link from Kevin) makes a solid point about tagging. It’s a difficult concept when your tagging text.

I think in many situations using solid user research to develop a taxonomy is much more efficient than relying on user supplied meta data about the content.

But it depends of course on the nature of the content. If it’s 100% user supplied then it can have more weight...after all you have no idea what the content IS.

I think tags are a good idea and we’ve seen some very good examples of how they can be used well, but all the hullabaloo about them far outweighs their actual usefulness.

On Flickr, for example, the reason tags work is because there are so many people doing it that some generally-useful patterns emerge, and because the individual content units aren’t particularly important by themselves.  On a smaller site (e.g., a corporate intranet), that wouldn’t necessarily be the case, because much-needed documents might be poorly tagged by a well-meaning employee (e.g., “tpsreports” instead of “tps").

On projects like that where it’s important that individual content pieces not be lost, and where the audience and content quantity isn’t large enough for patterns to emerge despite widely-varying methodologies, a more formal approach would probably be more appropriate.

It’s hard to give more specific recommendations without knowing more about what you’re doing, but I’d strongly recommend that you check out faceted classification as an option.

Two things to consider.

Who is doing the tagging? If the content author does it then the folks aspect of folksonomy is mostly lost and James’ point about quantity is magnified.

Are the tags used as the only subsequent means of access (as is currently the case with delicious) or does the system support search as well (Yahoo MyWeb, RawSugar--where I work)?

I’m a PhD student currently studying this whole, messy, area of user-centred categorisation. I think James Archer summed it up well. Folksonomies work when lots of people find that tagging benefits them each individually. Then you can get interesting/useful information from the sheer volume of people tagging for their own benefit. You can then do wonderful things with statistics and probabilities and whatnot.

However, with corporate intranets and knowledge management systems, etc., you often don’t have enough users to make free-tagging useful. You could however, allow tagging from a restricted vocabulary to prevent the synonym-tagging problem (eg. what one person calls ‘tps’ another calls ‘tpsreport’). The tricky thing is getting this restricted vocabulary to match the vocabulary that the users actually use, rather than what the designer thinks the users use, or even what the managers think the users use. Some writers suggest that you allow free-tagging initially, then come back and analyse the data later, and consolidate the vocabulary into a finite list of allowable tags/categories. This may not be practical in all situations however.

Bill’s point is good: who’s doing the tagging? One thing that annoys me about folksonomy systems is when I’m not presented (as a user) with a list of tags to choose from (I must come up with my own). I think someone mentioned that above.

I’m thinking about using tags for my next revision of my portfolio. This is because many of my projects have multiple media types or roles. BUT, I’m also going to apply categories to the “entries” - so conceivably users could navigate that way, as well. There’s also the search engine to consider - different users browse differently.

I agree that there should always be multiple paths to the information.

The librarian within me (and I am actually a librarian) wants to say that tagging will never be as useful as a controlled vocabulary. Yet, I think that user generated descriptors will evolve thru communities such as technorati and Flickr as these systems put more thought into this area. The tag-cloud is a good visualization technique. I think that there are going to be some big developments in the UI for tagging in the coming year. And I must say that I’m noticing more new traffic on my blogs coming via tags than thru search engines. There’s an interesting discussion on tagging taking place at tagsonomy.com

Considering how early we are in the use of tags there’s no doubt that better UIs will be found to express them and better ways to determine what tags to show at given points in a search path. That’s actually something our company is working hard on at the moment. I think clouds are interesting but when databases get large enough to cover a meaningful portion of the web the quantity of tags may overwhelm their usefulness.

My main problem with tagging is that it ‘gives up’ the attempt to provide a universally understood glossary. Multiple different ways to categorise the same ideas occur hence someone searching on just one of them will return a subset of data that doesn’t match what they want to find but matches posts by people who describe things in the same way. An upper ontology, however innaccurate, will at least provide a consistent (if flawed) way of categorising information. Perhaps the tags can be collected and assigned to a hierarchical categorisation structure and synonym lookup information provided, but it would seem to me that getting a simple initial category hierarchy, although more difficult and controversial, would provide better information classification as time progresses (let the users have some influence on the category hierarchies / or faceted classification schemes maybe)

have you looked at wikipaedia tagging? Use the wikipaedia page id’s as your tags… there are very strong definitions for what each ‘term’ means. For instance if I wanted to tag my writings on the python language I would use Python_programming_language as my wikipaedia tag (so it’s definitely not pets or monty).

I think that the issue of usage will come out in the wash over time as system developers learn to use the data to find connections. There will also be better ways of presenting existing tags so they can be reused rather than retyped. To some degree our system does both already though we’re not claiming anything mroe than a start in the right direction yet.

Mark B.: Any thoughts of how one might apply folksonomy approaches to Expression Engine? I’ve been checking out the Keywords plug-in, and I wonder if you could create a UI for pre-set tags (like Tim mentioned above).

I agree that letting users tag however they wish is likely to cause confusion for others searching.

Allan, I’m not an ExpressionEngine user but WordPress (my preference) has several tagging plugins available and you might look at those for ideascraping.

So, the thing now is to come up with a suitable UI which shows the relationships between the journal entries and the ?of interest? links.

That’s an interesting task; what might that look like? A list of tags at the end of a post ("design, typography, Macs") that link to a link list category page? Do I understand your use case properly?

I (and others) are also curious to see how you might solve this in an EE context. There’s a dearth of quality tutorials for that deep CMS, and one reason many from that community come here!

[J. Sinclair] However, with corporate intranets and knowledge management systems, etc., you often don?t have enough users to make free-tagging useful. You could however, allow tagging from a restricted vocabulary to prevent the synonym-tagging problem (eg. what one person calls ?tps? another calls ?tpsreport?). The tricky thing is getting this restricted vocabulary to match the vocabulary that the users actually use, rather than what the designer thinks the users use, or even what the managers think the users use. Some writers suggest that you allow free-tagging initially, then come back and analyse the data later, and consolidate the vocabulary into a finite list of allowable tags/categories. This may not be practical in all situations however.

Idea: what if you let users submit a ‘tag request’ in this scenario? If you had a list that was ‘good enough’ for most uses, then let users submit requests for the arcane ones (’tpsreport’) to a human ‘tag curator’ for consideration.

Hi,

Prompted by this discussion, I’m trying a slightly different approach based on a combination of hierarchy and folksonomy (or plain metadata). Here are the steps I’m building in : (quick glossary - hcat = hierarchical category, kw = folksonomy tag or keyword)

1) scan post for words that match hcats and show them with tick boxes for selection
2) let the user add their own kws and scan them for related hcats
3) let the user tick appropriate hcats
4) allow the user the option to (i) map any remaining kws to categories (synonyms) or (ii) place the kw in hierarchy (possibly moderated)
5) store the remaining kws in a ‘metadata’ field.

step 5 can be used if people later wish to add a new category and find related items (e.g. I add Cornwall as a kw and a later users decides to create a real category for it and searches in meta data for it).

This may seem long winded but it allows people to just add keywords if they like but it helps them to add (and start to create) hierarchical categories.

Any thoughts would be appreciated.

Thanks Mark,

I did intend for the ticking to be done once the article is finished so the tick boxes would appear at the bottom of the text. This could be done whilst the article was being written (but not efficiently with a large ontology).

Tim

This is something I am strugling with as well, and the comments many people have made already are helpful.

It sounds like this may be a bit basic for some (it goes ‘back’ as far as hierarchies), but I found the following article a good start: http://wiki.osafoundation.org/bin/view/Journal/HierarchyVersusFacetsVersusTags