Lazy Tagger

I’d see myself as a lazy tagger. I rarely think of new tags opting to fit a resource into either one of my own existing tag categories (with citeulike and del.icio.us), using tags entered by other users, or using tags that are suggested to me. Is this because I am lazy and can’t be bothered to think of new tags, or is it more likely to be an issue of time and how I like my internet experience to be as quick as possible and for my actions to be as intuitive and automated as possible, before I flit off to the next task. This behaviour is measured as cognitive cost. Tagging is so popular because of its low cognitive cost to users, but is this having an effect on the quality of tags that tagging systems are generating? Can folksonomies be accurate, if like me, users are only tagging in agreement with the first user who tagged the resource?

Research suggests (Golder & Huberman, Catutto – I’ll post a reference list!) that high frequency tags, namely tags created from the process described above are semantically lower quality, using a basic specificity level of vocabulary than lower frequency tags, which use more specific, subordinate level vocabulary. This has implications for the use of tag data for navigation, classification, categorisation and descriptive text.

And I’m guilty of it. My tag data is probably of no use to anyone, it is barely useful to me. I have this need to keep tags in categories and will even opt to not use a really descriptive tag because it will be on its own and not belong to my ordered groups. I am consciously starting to stop myself doing this – I want to see lots more 1’s in my del.icio.us tag frequencies so I can plot a long tail.

Incentives to tag descriptively

I actually wrote this 10 April 2007 and for some reason didn’t post it. I’m saying this because i’ve now wrote a similar themed post but after a couple more weeks of reading it is more informed. so i’m posting both.

When I first started tagging on del.icio.us I tagged very selfishly, tagging everything MSC-project. This is only useful for me to sort my bookmarks by what is relevant to my project, it is of no use to any other del.icio.us user unless they happen to know what my project is about. As it turns out, I only use del.icio.us to manage the bookmarks for my research project, so the tag is redundant and absolutely no use as a filter as it would retrieve all my bookmarks. Well up to a certain date, as I have now realised the uselessness of that tag and no longer use it. I am also a culprit of not tagging very descriptively, I went through a phase of tagging everything web2.0 until I realised how every bookmark will be categorised as web2.0 and ultimately it is not useful in sorting my bookmarks. I have tried to tag more relevant to the resource, however I have discovered I am a lazy tagger and will only really use the suggested tags that del.icio.us kindly presents to me when I am adding a new bookmark. Sometimes I’ll add a new tag if I don’t see one that accurately describes the resource.

 

As for other systems, last fm, you tube, I can’t really be bothered to tag anything, why should I ? There is no benefit to me to tag. This behaviour is common in taggers. You only have to look at he tags entered in last fm and you tube to see how un descriptive the tags are and how selfish the tags can be. Typical examples include Rock, or Indie are the most used tags on last fm, Seen live is another common one, which is an example of a selfish tag as it is not of benefit to anyone else who hasn’t seen the artist live.

 

I have found these traits in my own tagging practice and in looking at how others tag, it is also a new research area, with researchers trying to determin what incentives encourage users to tag. (Marlow et al., Zollers) It is this research that has inspired my own project. What incentives will encourage users to tag more descriptively? I am going to take lead from Luis van Ahn of ESP game fame by creating a tagging experiment to see if by making the tagging if internet video resources fun, more descriptive tags can be collected.

 

nervous tick

My stress levels for this project are already reaching the levels they get to toward the end of a module, when there’s a couple of weeks left until a deadline and i still have practically the whole assignment left to do.  I always got it done and did it well too.  I’m seeing my literature review as a deadline and i need to get that done before moving on to anything else.  I have developed the nervous tick in my right eye that comes the week before a deadline normally.  I have 5 months before my project has to be submitted i hope i don’t have a nervous tick for that whole time i’ll have a full head tick by the end then i think.  I’ll feel loads better when i get the lit review done, i can’t believe how much grief it’s causing me.  Maybe i need to stop writing stupid blog entries and actually start doing some work, that may help.

Literature Review

I promise myself i am actually going to write some thoughts regarding the research i’m doing rather than project management ramblings.  My mind is all fuzzy at the minute, i have spent the week going round in circles with my research questions, i have too many.  I can’t possibly expect my little MSC project to answer them all, so i am going to have to relegate some to the future work section of my report.  I will however discuss them through literature review, because they came from reading papers after all.  When i am clear i will write a bit in here too.  But i have now managed to get one clear (ish) research question for my whole project, which is how to create incentives to encourage users to tag internet video resources more descriptively?

So literature review, i need a plan, i am procrastinating a bit on it, i keep reading and reading, falling asleep through the reading and then reading a bit more, but it’s getting me nowhere fast.  I have decided to take a suggestion from my project handbook and create a taxonomy for the 15 odd v.important papers i have in order to highlight which papers need to be discussed in what section and how they can be compared and contrasted.  In theory this is a great idea, in practise i need to now be clear what sections are going to go into my literature review and what questions i am expecting my literature review to raise and/or answer.

I know the knowledge is there it’s just in a massive jumble in my head and i can’t get it out clearly into structure and words.  Once i do i’m actually going to write something in this blog that hopefully sounds intelligent and will give some weight to the reasons i’m doing this project and why people should help me by playing my game.

Progress Update

Since my last post, i have designed a suitably complicated database and normalised all the tables, drawn an entity relationship diagram, then went back to finalise the requirements analysis and discovered i’d missed bits out of the database. There’s a reason they teach you to do things in a certain order. I’m being very disciplined here, because it is so tempting to go straight in and start programming the game without any formal design steps, but as i need them for the report i am doing them now rather than at the end. And to be honest, the steps are very helpful, it is better to discover a completely overlooked table at the start of the project than halfway through when a lot of programming has already been done. I have only used design and development methods for software in university modules, i’ve never applied them to a website, or any of my own work, but the processes are the same. I am planning to follow a waterfall lifecycle model, although i have already demonstrated that a more iterative approach will be taken as i have already flipped between design and requirements analysis.

I also did some important work buying a domain name www.videotag.co.uk for the game and i designed a logo. I am quite proud of the logo, i found the Atari font for free on the Internet, from here. They also had a yahoo one which i messed with but it looked crap. Although the Atari font says arcade game and i’m not making an arcade game, i like the idea of creating a simple retro feel to the design for what is effectively a new technology. I read this site that categorises web 2 logos into softies, futurists and classics and guesses what font they used. I think mine falls somewhere between a softie and a futurist, but it looks web2 and that was my aim.

So now i have the all important logo design out of the way i need to concentrate on the literature review. I went to a doctoral forum last week which was really helpful in focusing my research, it highlighted all the holes i have, which i can basically fill by completing the literature review and doing a bit more reading. so that is my next goal. As you can see i am distracting myself well by waffling away in this blog instead of getting down to some reading!