Snowdrop or snowdrops?

I have now read the article on Folksonomies: Tidying up tags by Marieke Guy and Emma Tonkin, one of the tasks in the Multiliteracies sessions (Electronic Village Online 2009). It has set me thinking about my own – limited and very basic – experiences with tagging. I have tried to tag blog posts in Blogger and WordPress, photos in Flickr, music in LastFm, and videos in YouTube, for different purposes and with varying success.

Blogs
I have now put 144 posts up in my personal Micalet blog on Blogger, over a period of 22 months. I appear to have used 492 tags in all, giving an average of 3.4 tags per post. There are 239 different tags; the most used ones appear 29 and 28 times, only 9 reach double figures, and 178 of them have been used only once. This seems to confirm the pattern found by Guy and Tonkin – that’s a relief, at least I’m normal!

The most used tags are for the places I live in, and that I visit on holiday, and for ‘Music’ and ‘English music’ – all very general; amongst the once-only tags are many that are quite specific, such as names of musicians, or terms such as ‘spider’, ‘snowman’ or ‘wheel clamp’ – I hope I don’t have to use that one again! The pattern appears to be one or two general tags, plus one or two specific ones, per post. I don’t think this was a conscious strategy, it just seems to have evolved, but in the light of the article it appears to be a useful one.

According to Sitemeter, I have had 195 visits since I started counting in November; I haven’t checked where these have come from, but hardly any of them have left a trace, as there are very few comments on the posts. Whether the tags have contributed to the number of visits I have no idea, and I’m not sure if there’s any way of knowing.

So what use are tags in blog posts to me?

Well, I have come across one very useful function. Amongst the widgets I have set up is a tag list (this is where I got the data above from). If for any reason I wish to gather together posts relevant to a particular topic, I can click on that tag in the list, and Blogger composes, on the fly, a new page with only posts that include that tag. So, if by any chance you’d like to read what I’ve put up on English music, Blogger gives me the page: http://blog.michalska.net/search/label/English%20music , and I can quote the url to you.

This has proved really useful with a Blogger site – San Miguel Bajo – which I set up to house materials for students who had difficulties accessing our University’s VLE. I just put things up as they come, and use tags for each course so that I can point students towards the material relevant to them. For instance, the course tag SP201 furnishes a page for my advanced class.

Photos
I have just over 1000 photos on Flickr, and according to my Flickr statistics, I appear to have tagged all bar a handful. My usual approach is to give all my photos ‘keywords’ (ie, tags) as soon as I put them into iPhoto; then when I upload them to Flickr, the tags are retained. Within Flickr I put most photos into Sets, and you can then easily edit the tags for all the photos in any given set.

Once again I have no idea how effective these tags have been. Over 60% of the photos have been viewed, that sounds a lot to me, but I don’t know what other people get; neither do I know what part tags played in attracting these visits. Mostly there are a handful of visits per day, occasionally up to 20 or so. Recently there have been a couple of exceptional days – 121 on Jan 26, and 192 on Jan 21 – I have no idea why, nor who, nor where from, nor if it’s in any way tag-related.

Very occasionally, I do get a message saying someone likes a photo, and would like to include it in a Flickr Group – which is what happened with the snowdrops, up there at the top of this post. The only way I can imagine they came across this picture is via the tag: snowdrops.

[As I write this, somebody I have never heard of has just added me as a contact in Flickr. This person already has 243 contacts, and subscribes to over 500 groups (I pasted the list into Excel to count!). Why did he pick on me? I think I’ll ask him – you never know, we may have something in common . . . ]

[5 minutes later: He’s beaten me to it – he’s asked to join the only Flickr Group I run, Dansez Français, which so far has only me and one other in it. We could do with a bit of a boost, so I’ll accept him . . . ]

Music
I have managed to get LastFm to list most of the contents of my iTunes, and have tagged a fair proportion. This has to be done within LastFm, as iTunes – unlike iPhoto – has a very underdeveloped (ie, useless) tagging system. Here I come head on up against several of the issues described by Guy and Tonkin. You may have your own ideas on how to classify a particular artist, or piece of music, but LastFm has a well-developed folksonomic (?) system, and first offers you the tags that others have already used. These may differ slightly from yours, or be completely different. So do you use theirs, which you may not be quite comfortable with – they may miss an aspect you feel is significant – or do you insist on adding your own, which puts you in the field of unique tags again?

And how do you handle other people’s errors? Tagging something ‘flamenco’ when it comes from the Basque Country (oh-oh!). You know it’s wrong, but someone’s already done it, and others may have followed their classification, and the size of the offending tag grows and it becomes more prominent and hence ‘authoritative’. You can’t change other people’s tags, you can only add your own. The Folksonomy is supposed to exemplify the ‘wisdom of the crowd’. So what can you do about the ignorance of the crowd?

I have also tried to use the tagging system to create ‘radio stations’ of various descriptions, which I would like to contain just selections from my own music collection, which I can then offer to friends, or possibly students. However LastFm is a social network, and insists on being social and including tracks from elsewhere on the site. This puts my ‘station’ at the mercy of other people’s errors, as outlined above, as well as more fundamentally their classification boundaries, which may not coincide with mine, and their taste, which will certainly differ. I may like track A by a certain artist, but not track B – but I will be fed anything anyone has tagged with the same tag as I have used. I am thereby subjected to the tyranny of the crowd.

Videos
YouTube’s response to tags is a law unto itself. The last couple of videos I uploaded were films of different versions of the bourrée, a traditional dance from Central France, so I thought I’d try looking for ‘bourrée’ on YouTube, to see what else was there. You will already be anticipating some of the issues – tags (or titles, or descriptions – YouTube searches on them all) using a language other than English, words that use a diacritic (é in this case) – I was expecting these to cause problems, however they didn’t seem to bother YouTube. However I didn’t learn much about French dance from the first page of hits, using the ‘relevance’ filter – 19 of the 20 were classical music pieces, mostly movements or sections from Suites by JS Bach, plus a piece by Jethro Tull. Gorgeous music, for sure, but not what I was looking for. The only film purporting to be of the French dance was an excruciatingly poor example – I won’t go into why, just trust me on that.

So here are two more stumbling blocks: words used in more than one context, and content that doesn’t match up to the tag.

When I then changed the filter to ‘view count’, the surprise was even greater – 7 out of the 20 on page 1 are films of people in various states of inebriation – ‘bourré’ is also slang for ‘drunk’ in French. And not one single version of the dance.

However, when I start at one of my own films, YouTube seems to have a better idea of what I’m after, although it’s still not perfect. Try Bourrée à 4, à 2 temps, and check the Related Videos list on the right – the first 10 are spot on, after that it’s back to Jethro Tull and JS Bach.

So what’s happening? Maybe YouTube needs more information from me, given the obstacles it has to surmount. I could put a second tag in, but I would be second guessing how the original posters had tagged their uploads, and those that have used more than one tag are unlikely to coincide. And of course there will be some that haven’t tagged at all.

Judging by the number of people you see filming at workshops and festivals (I’m not the only one!), I would guess there are scores if not hundreds of bourrées out there. Somewhere.

Snowdrop or snowdrops?
Amongst the many issues raised by Guy and Tonkin is whether you should use the singular form or the plural in tags. As I understand it, librarians prefer the singular: all photos containing one or more snowdrops should be tagged ‘snowdrop’. My snowdrop set has 7 pictures, in one of which I somehow ended up with just one snowdrop – but I decided to tag them all with the plural form nevertheless, reasoning that if I were looking for pictures of snowdrops, that is what I’d search for: snowdrops. I’m not sure what difference that makes to Flickr, but I do know that you can now see 1866 photos of the little lovelies at the Ghiocei || Snowdrops Group Pool, contributed by over 300 people from many different countries. So despite its imperfections, the world of Folksonomies does have something going for it.