I don’t think I have ever talked about data mining on this site, but it is an extremely powerful concept that you will become quite familiar with soon. Well, in about five years or so.
The actual definition of data mining, as per wikipedia, is the process of automatically searching large volumes of data for patterns using association rules (aka algorithms). When you get right down to it, data mining is exactly what Google does every time you type in a search string. I will get into more data mining in a bit.
Now, on to Swivel. Swivel.com allows users to upload data, and then the it arranges that data graphically, while making it searchable. Other users can then mix and mash whatever information they are interested in. The info is displayed in graphs to create visual relationships. The express purpose of the site is to spot trends through all sorts of data, uploaded by random people.
For example, World of Warcraft and Everquest II:
Truly, the value of the site lies with the information that users are uploading and the relationships that can be built between the numbers. In order to carry out it’s mission well, it needs to have good, hard nosed data. Who is to say that the information being submitted is of any quality whatsoever? After all, this is the Internet!
Therein lies the problem. For as powerful as data mining is, you ABSOLUTELY NEED TO HAVE GOOD DATA. If the quality of whatever you are filtering through is terrible, then the output will be terrible. I have some horror stories concerning data mining, and I will be the first one to say that 80% of my time was spent ‘cleaning the data,’ or making sure that it was good.
On the other hand, if you know how to design a good database or a good set of verification methods, the quality rises exponentially. And the uses are extraordinary. For example, Walmart has a data mining program that can tell what demographic bought what at where in a store. For this reason alone a couple years ago, they moved baby diapers next to the beer and put one of them on sale every Friday night. (A sad world we live in, I know. But their sales increased!)
What is the old adage, “Garbage in – garbage out?” I think Swivel could do some amazing things. I just hope that they have a system in place that filters all of the data.
No related posts.
[...] data for testing a new algorithm or applying a developed methodology. Concerning the latest one, JDSBlog proposes a possible solution through Swivel.com, a website allowing users to upload their data. The [...]
Low Priced Humidifier
Buy Humidifiers at Wholesale Prices Free Shipping on Orders Over $149!