Some think a database is just a place to store things. Others think of it more as a library wherein information is stored and then creatively retrieved, creating new metrics and ways of looking at the data. This is called Analysis Services to augment business intelligence.
To compare the two, let's look at an example to better visualize why one is superior to the other. Imagine you are holding a thick book in your hands. If you want to find every reference to -ants- in that book, you will be required to go through each page, reading every word, and marking each instance of the word -ants.- This is a linear search - it goes from front to back and finds only what you specify and nothing more.
Were you to do the same thing with the same book looking for -ants- and anything related (-bug-, -ant- in the singular, -insect,- etc.) you would go through the book in the same way, marking everything related to ants. It takes a little more time, but you have more information when you're done. Rather than just referencing -ants- in the book, you've found everything that includes -ant,- -bug,- -insect- and so forth. Using this, you not only know where it references -ants- but you have contextual references as well, giving you more depth.
Now imagine that your computer was only capable of linear searching, even online. Instead of Google returning relevant results based on your general subject matter, it returns only what you type. So if you Google -ants- it will return only things that specifically have the word -ants- in them. Instead of -ant,- -insect,- and so on. Now imagine that Google had to go through all of its servers (millions of them) one at a time, front to back, for each search.
Not very efficient.
Analysis Services work in multiple dimensions, which means that they pull data from several places at once rather than one at a time. So imagine that all of the information in Google is stored on a two-dimensions, flat plane - a giant piece of paper. To find things, Google could either start at one end and go through the whole thing, inch by inch, or it could go multi-D. Then it would look from above, below, both sides, the middle, etc. Much faster.
Your businesses servers could work like that. Most likely you are using SQL Server for data storage - it's the most common. If so, Analysis Services are easily integrated into SQL Server 2008 and R2 machines.
Although Analysis Services are complex, they are extremely useful. They give the business many ways to not only retrieve, but look at and analyze data. Analysis can be deeply complex or relatively simple, but are always value-added over the linear approach.
To compare the two, let's look at an example to better visualize why one is superior to the other. Imagine you are holding a thick book in your hands. If you want to find every reference to -ants- in that book, you will be required to go through each page, reading every word, and marking each instance of the word -ants.- This is a linear search - it goes from front to back and finds only what you specify and nothing more.
Were you to do the same thing with the same book looking for -ants- and anything related (-bug-, -ant- in the singular, -insect,- etc.) you would go through the book in the same way, marking everything related to ants. It takes a little more time, but you have more information when you're done. Rather than just referencing -ants- in the book, you've found everything that includes -ant,- -bug,- -insect- and so forth. Using this, you not only know where it references -ants- but you have contextual references as well, giving you more depth.
Now imagine that your computer was only capable of linear searching, even online. Instead of Google returning relevant results based on your general subject matter, it returns only what you type. So if you Google -ants- it will return only things that specifically have the word -ants- in them. Instead of -ant,- -insect,- and so on. Now imagine that Google had to go through all of its servers (millions of them) one at a time, front to back, for each search.
Not very efficient.
Analysis Services work in multiple dimensions, which means that they pull data from several places at once rather than one at a time. So imagine that all of the information in Google is stored on a two-dimensions, flat plane - a giant piece of paper. To find things, Google could either start at one end and go through the whole thing, inch by inch, or it could go multi-D. Then it would look from above, below, both sides, the middle, etc. Much faster.
Your businesses servers could work like that. Most likely you are using SQL Server for data storage - it's the most common. If so, Analysis Services are easily integrated into SQL Server 2008 and R2 machines.
Although Analysis Services are complex, they are extremely useful. They give the business many ways to not only retrieve, but look at and analyze data. Analysis can be deeply complex or relatively simple, but are always value-added over the linear approach.
No comments :
Post a Comment