Week 11- Data Visualisation

Well, this was our last lecture, sad no? But it did leave me personally with a lot to think about. Data Visualisation should well be everywhere! and they are and I can't help but feel like shouting "Why wasn't I informed?!" cause you can do so much and connect information that well frankly might not be connected! Say for instance that Colea example people were getting sick they did a data visualisation and where able to deduce that it was a particular pump on a particular street corner that was the source of the trouble!

I thought that the idea of inbuilt visual wetware was a pretty good way of looking at it. Shapes, scale, colour, movement and position. These are the elements that effect the way we interpret information and were pretty good at it.

So by my immediate enthusiasm for dataviz I'm sure its becoming apparent what I intend to do! But the question is what kind of data am I going to be going to look at? My itunes library! and you must be like 'why would we wanna know about that?' but I'm going to track my music listening back as far as I can and then look for correlations between the music and my life and events that have influence what and how much I was listening to and if that's not good enough then I'm going to add in a friend's library for comparison and see how our listening habits and tastes have deviated from each other. Fun no? So I will be looking at play counts, date of import, date last listened and see if we can find any music experiencing the long tail effect and what is really just a fad.

Week 10- Tute

So, been thinking about the idea that I was ranting about the other day- alternative to tag clouds and I have further developed it with a couple guys in my tutorial. So essentially if you remember this started off as an alternative to tag clouds. However, the site design has now become a service that will connect designers to companies that may need a fresh design for their business or a band that need a new t-shirt design. It's all about connections.

So what we have suggested is that we have a cloud graphic which you can search a tag for instance water. This will then bring up nodes of words that are related to water such as waves- then this allows you to look at designs or graphics that are wave related. From there you can click on a image and this will bring the image up and blackout the rest of the screen. You will have the graphic, a brief synopsis from the artist what it is about,  a method of contact and other works from the artist which you can dive into so sand graphic would then allow you to see his and then others designs involving sand.

With this we would like to offer user accounts for artists in doing so we could set up a chat system so that consumers could have live chats with artist about work they would like them to be involved in.

Technologies that this could involve- Sea Dragon so we have that smooth picture diving capability, APIs so that artists won't have to re-upload work to our server, we would probably rent out server space off-site for user information. Possibly page rank system for the popularity of particular designs and incorporating consumer feedback.

Possibly drawbacks- Bandwidth always with the bandwidth!

Ways to monetize- commissions so we have a small commission for providing the medium and contact between artist/designer and consumer.

I do have a step by step images but I did them in Flash so will have to post them later as I don't have flash on this system =)

Week 10 Lecture- Multiplicity

This weeks lecture was looking at a couple different things that all really tie together.
-Sampling
-Solutions to Insufficient Bandwidth
- Scale
-Iteration
-Collaborative Filter
-Crowd Sourcing- The wisdom of the crowd.

Rather than go through each thing individually this week I thought I would do just a quick overview and then talk about some articles I've been reading that caught my eye on the web.

Essentially, in the lecture we were looking at the wisdom of the crowd. The idea that the collective can do a better job than a single individual.
The example that Michael gave in the lecture was about the weight of the cow at a fair and the person who guessed closest won the cow. But what was cool was the average of all the answers was closer than the person who won. Now you might wonder what does this tell us? It tells us that the idea 'collective intelligence' is not restricted to the web and that a collective can answer better than an individual because each person brings their own knowledge to the table until you have an ultimate table of amazing intelligence or until you run out of chairs or until they realise the need to send someone to buy more chairs.

Iteration so doing a process over and over till you reach a result or destination. Now this is done with computers with music and video which Michael mentioned in the lecture. The computer reconstructs the audio or video till humans cannot tell the difference between the original and the copy.  This ties into solution to insufficient bandwidth- we can do fewer samples of the mp3 and it would take up less bandwidth. The other option is we develop a better method of compression.

So Michael quickly touched on Twitter and the scale it took off means that the Twitter guys needed to develop a way of monetizing it. If you haven't seen it Michael put an article up on moodle and though it was predictable, I must say that I'm a little disappointed with their solution (they have to make money I guess, least its a new way to present ads). The people at Twitter propose to put advertisements in your Twitter feed, however these ads will be directed at you specifically depending on key words that companies can buy. So if Starbucks wanted in on the action which I'm sure they will they might buy 'Starbucks' and 'coffee' for instance. Then depending on how often you tweeted using these keywords Starbucks tweets would come up in your feed to see. Things like discounts and offers and things. Twitter is currently charging companies for every 1000 tweets until they finalise their business model.

Personally I don't use twitter, don't really care what your doing right now. But its something that has taken off so hats off to the technology.

Another cool site I checked out earlier this week was 99 Designs. The idea behind that is you can come to this site and pay the site to run a design competition on your behalf. Then designers submit their designs and at the end of the competition the designer I imagine wins the contract and you get buy the rights to their designs. Really opens up the design community.

Anyway that's enough banter for today.
Lastly really can't wait to get into photosynth that's really cool stuff, internet has been shaped 4 more days sigh....

Week 9 Lecture- Web 2.0

 I find it hard to say without sounding geeky but I found the idea of Web 2.0 and its future relying on us really exciting. I read the web 2.0 and long tail articles before the lecture and could instantly see the community that the web has become. Up until this point I had taken this as a granted or given but the article web 2.0 really put it all into perspective. That collective intelligence on the web has resulted into things such as the long tail model is phenomenal. After reading it you can really get an appreciation for what people at the forefront of web 2.0 are trying to do.

I thought that the page rank system that Google implemented was interesting. Which pretty much catapulted them to internet glory in which they were able to start establishing their empire! Now I use the term empire due to a little segment I saw on the ‘Hungry Beast’ which you guys might like- http://www.youtube.com/watch?v=R7yfV6RzE30 unfortunately it’s not available on the site anymore but I still recommend checking it out!

The way that AJAX was implemented it the webpage Twistori I thought was really impressive and it had me thinking about possible webpage designs that could implement it in a similar way. My immediate thoughts were about tag clouds, not that I have a problem with tag clouds just they can be really ugly sometimes depending on the colours the designer picked. It can sometimes just make a page look a bit disorganised. So working on the same idea that Twistori put together, why not use it in a similar way for tag clouds. We create a search bar in a cloud graphic then when a user searches a tag AJAX does its thing and we have links to articles that are related to the tag fall out of the cloud down the page. When the user clicks on a tag as it passes we have the article pop up in a blackout window. User can then return to continue searching through articles.

We have two options with what happens with the articles that are still moving. It could just be a continuous cycle of a few articles or could we program the AJAX to freeze the feed of tags while the user is reading?

Anyone have any thoughts?