Someone once said to me that if you do something once, it's an accident. Do it twice and it's a coincidence. Do it three or more times and that's just the way you're living. The underlying message is that if you repeat something enough, then the patterns of use start to tell their own story. Your repeated activity starts to build up into a pattern of use and looking at those patterns can often give insights into the activity that are not apparent by looking at the individual instances of the activity.
This idea of allowing data to "rest where it lays" and deriving insights from it is essentially the idea behind tag clouds, whose patterns reflect repeated use of words, tags, keywords or ideas. If you look at someone's Delicious tag cloud and see the patterns emerging in the form of highlighted, emphasised words, then you see a clear indication of what interests that person. The more they bookmark using tags, the more evident their interests. The numbers don't lie when there are enough of them.
if you aggregate enough tag clouds you start to get an insight into the "patterns of the patterns" - you see not just the interests of individuals emerging, but the interests of the group. This is the whole notion of a folksonomy, and it taps into the fascinating concept of the "wisdom of the crowds". Data, especially when you have enough of it to form reliable patterns, starts to become very interesting.
In the same spirit, I was a little intriugued by a twitter app I saw today, called TweetPsych. TweetPsych looks at the contents of your last 1000 messages on Twitter, analyses the words you use and the way your sentences are constructed, and tries to draw conclusions about what you do, what interests you, and what sort of person you might be - psychologically speaking. I've no idea how accurate it might be, but it's an interesting idea. I'll be honest and admit to you that I have absolutely no idea what they really mean, but here's my results anyway... http://tweetpsych.com/?name=betchaboy.
Regardless of whether TweetPsych is accurate and up to scratch just yet or not, I think it signals an interesting development in what is sure to become a much bigger deal. The notion that some level of machine intelligence can be derived from an analysis of massive amounts of our online footprints. We are all leaving massive amounts of data behind us as we trawl around the Net, and somewhere in that trail of data there are machines piecing together an accurate picture of us... what we like, where we go on holidays, who we talk to, what our preferences are, and so on. It's not a new idea - Google's entire advertising strategy is based on the concept of knowing more and more about you - but seeing TweetPsych's attempt at psychoanalysing me from these 140 character snippets of my thoughts just threw it into a new light.
Let's just hope that this data can be put to use in positive, creative ways that help enhance our lives.
I think this makes a valid point. SOmetimes it seems as though we don't really think about the trails we leave behind us, but every thing we do is one more step leading back to us. I think it is great that there is technology to do that but like you said, let's just hope it is used for a good cause because you never know when the wrong person might find one of your footprints.
ReplyDeleteWe are what we tweet and we is what we blog. I agree Chris. We are indeed leaving footprints. I wonder about sites like MySpace and Facebook. There is seemingly a lack of control and ownership there. Blogs such as your own enable you to exercise more control. With Facebook and MySpace you can be at the mercy of your 'friends'.
ReplyDeleteYet, let me get to the point. I did the TweetPsych test and I am rather happy to say that I am like John Cleese. http://tweetpsych.com/?name=john_larkin
That has made my day.
Cheers
John
Hi John,
ReplyDeleteI agree completely about the personal blog vs Myspace/Facebook thing. Interesting that my 13 year old daughter started a Blogger blog in preference to a Facebook account like all her friends. Her reasoning? She said she didn't want a page that other people had more control of than her. She prefered a blog because she could decide what went on it. Clever girl that daughter of mine!
John Cleese eh? Can you do the silly walk? ;-)
To tweet or not to tweet, that is the question. (What would ol’ Will of the Shake say if he were alive today? Why use Twitter? I find it fascinating, primarily because I don’t feel the need to let everyone know what I’m doing or thinking. I have a Facebook page, a sort of hunters and collectors (not the band) anthology, and a MySpace page, that I use to promote my own music. Do they reveal anything about the real me? I would have to argue no……..
ReplyDeleteObviously an opinion / picture could be formed about me based on what those sites contain, but since I control that, it ensures that my real life remains just that.
On a lighter note, of themselves, patterns can become interesting, but in the end what do they really reveal.
If you were to look at my blogs, my tag clouds only reveal what I want to use for search bait and don’t indicate anything about my actual interests. I try not to repeat tags entered on posts, but damn, when I look at that tag cloud, there they are like ghosts that haunt, evidence of repeated usage. I sit back and review my own processes……..what kind of footprint am I leaving behind, will someone construct a fictional persona and ascribe it to me based on the bait I use? Do fish ascribe the same to fisherman?
What you could argue is that you are looking at patterns of communication rather than any data that could facilitate an in depth analysis of the personality. If I were to use Twitter, could valid inferences be drawn from my tweeting? I would be so tempted to make it completely fictional, and perhaps there is an element of the fictional in the reality of the process, after all don’t you twitter only what you want other people to know? That leaves the remainder of what is unsaid as the hidden person. I can’t gauge anything about Chris from his twitter trail. I can only see as laid track.
As the creator of TweetPsych emphasises in the disclaimer, it’s primarily for entertainment purposes. Which in itself begs the question; can we separate entertainment from reality? And do we believe our own fictions, even in the face of the primacy of the empirical world?
Am I leading you up the garden path?
Hi Gary,
ReplyDeleteI think that you and I probably use Twitter quite differently. And despite your claim that you only show "bait" in your online profile, I'm pretty sure that, over time, I could build up a much better idea of who you are and what your interests are than you suspect... if I took the time to study your Twitter feed, your blog, your Delicious links, your Facebook page, etc... I think the picture it leaves behind of who you are might be more descriptive than you let on.
Thanks for the comment.
Chris
Hi rewired,
ReplyDeleteAfter having read Dave Weinberger's "Everything is Miscellaneous" I have to say I completely disagree with your assertion that traditional hierarchical navigation can possibly be the right solution to coping with the amazing miscellany of our world. There are lots of obvious flaws with "library science" and almost no perfect examples of such systems doing a comprehensively good job of dealing with massively miscellaneous data. The Dewey Decimal System is a complete joke when you really look at how it works. Taxonomy systems like Carl Linnaeus's do a great job of categorising the obvious, but a pretty lousy job of dealing with the obscure, and the more random or miscellaneous the information the harder it is for traditional categorisation systems to deal with it.
The flaw in this argument is in thinking that tag clouds are primarily a navigation tool... they shouldn't really be used as that. They DO give interesting insights into massive amounts of data, and they CAN reveal interesting patterns that may or may not have been otherwise obvious, but they aren't really meant (in my opinion) to be a primary source of navigation. Tags ARE really good at adding hooks to data that might otherwise be hard for search engines to lock onto - pictures, links, videos, etc - and therefore they help make the unsearchable searchable. In your example of pictures of Taksim, tags would be an ideal way for search technologies to latch onto relatively obscure data.
Sites that rise briefly to popularity after getting Dugg might have a short term win, but they quickly fall back off the radar. The real point is that without the ability for the public to draw such sites out of the crowd, they would get lost in the blur and never get any visibility. The tags in this case give visibility, no matter how briefly, to things that people DO find interesting. If Eight Mile is interesting to people, why shouldn't it have its moment in the sun? It's not like Detroit is not available - search will still find things that people really need - but the folksonomy will expose things that are not so obvious.
Thought provoking comment, thanks for posting it.
Chris
Hi Chris
ReplyDeleteI have to apologize for not following net protocol and failing to correctly reference the quoted 'part post' from Jeffrey Zeldman. Had I done so the text would not have been attributed to me....hence ,"I have to say I completely disagree with your assertion that traditional hierarchical navigation can possibly be the right solution to coping with the amazing miscellany of our world."should be directed to Zeldman http://www.zeldman.com/daily/0505a.shtml.
Like you I do not necessarily agree with all of Zeldman's assertions, I simply placed the referenced text as a post to stimulate conversation on this very interesting subject. I did place the text in quotation marks and provide the web link to Zeldman's article, but perhaps I should have been clearer on this.
You could follow this up by referencing the full post of that was quoted in part......
Remove Forebrain and Serve: Tag Clouds II
Which is prefaced by
In “Tag Clouds are the New Mullets” (Daily Report, 19 April 2005), I claimed that the weighted tag clouds meme popularized by Flickr and Technorati was about to cross a permanent shame threshold because of overuse. My comment suggested that the only sin of tag clouds was popularity. But the problems with tag clouds run deeper.
All the best
Hi Chris
ReplyDeleteWell I don't use Twitter, nor Delicious, but after some research I've discovered maybe I should. However you are more than welcome to offer an analysis of me based on what you can find. But in the interim here is something a little more interesting and unbeleivably informative.
http://adage.com/agencynews/article?article_id=137724
Cheers
Gary