Data mining is simply about getting all the data your organisation has access to and using it intelligently.
In this age of computer gadgets, companies can collect data from you, your social media, your mobile phone, etc..
The first time I learned about Data mining was on a company paid course (one of my previous employers) and the issue was “we will be collecting a lot of information from customers and potential customers so it must be used to the benefit of the company”.
Where Uganda falls behind is not realising that they already have much information from the bibiriti they have given people and so many Ugandans expose a lot on FBzero.
For data mining to work to your advantage, you need a strong database in place. Our most recent elections show that Uganda was not paying attention to databases. You can refer to my previous article about the Biometric Voter Verification. http://nangalama.blogspot.ca/2016/02/biometric-voting-for-uganda.html
There are some phrases you will here about data mining.
Data warehousing. This is simply a method of collecting data, putting it in databases like Oracle, DB2, SAP, MySQL, etc.. Once you put it in, then you have to look into it with an intelligent eye to generate reports which can make sense for profit.
CIF (Corporate Information Factory). This includes collecting all the data you can from your operations, storing it in one place (databases) and then generating daily, weekly or month reports so that the company can know about their sales and losses.
ETL (data Extraction, Transiformation and Loading). This simply means that once you have your data in a warehouse (read data base so you need a DBA), you can grab it from any database or software transformation, transform it into any language that can read and load it into the destination. Most people who work with data know that we can take a flat file or comma delimited file and move it into SQL. SQL stands for Structured Query Language and this is what most if not all databases use. Databases are the foundation of the reports and forms you see when you use applications.
Archiving. Data archiving is crucial. You can collect all the data you want, transform it, put it in databases and create canned reports from it. Till your warehouse crashes and you have no data. Generally most companies archive (backup) their data nightly and in various places in case the building blows up. Read about cloud storage here. http://nangalama.blogspot.ca/2015/02/data-storage-basics-uganda-pay.html
Most managers, project managers and supervisors work with reports. So do the auditors. And sometimes the investors. These people depend on custom generated reports or canned reports. Custom generated reports are reports we can create on the fly. We have come a long way in technology that we can now just ask the user to specify what they need and then we generate the report immediately. Canned reports are ones we have prepared previously and provide a menu for people to specify their needs (eg. Uganda, Kenya, Tanzania, sales, etc..) and those are canned reports.
I had one experience in Moncton Canada where I flew into a bank in panic because my purse had vanished (not the first time) and I had no ID. The customer rep looked at me and asked just for my name. I was amazed to get a new bank card in less than 5 minutes and asked her how she did it. She turned her computer screen to my face and the verification was “black, female, about 35yrs, smiles a lot..etc..” This is Data Mining. Everything you do is available for digs.
Sears is a well knowned Canadian company and also in USA. Sears was the pioneer of Data Mining. This was way before all the above things happened in technology. Sears can be found on Google but good luck finding out about their data mining. Many more Canadian and USA companies have followed this model.
You purchase something from Sears and you go to the check out. They ask for your postal code and your name and maybe not even your name. But the person behind that computer is writing in it “35yr old looking female, two young girls looking 5 and 2, bought jewelry, toys for the girls, was not accepting help from the sales people on the floor, kids like toys, etc..”
The next time you go into a Sears store, someone walks up to you to sell you that perfume brand you last bought, offers kids things from the Disney section, knows to keep her distance, etc… So you see, data mining can make you profitable. You can use it to know your customers and their next customers in that these little monsters grow up and return to the same place they got their Disney toys from.
Information you could use.
Martha Leah Nangalama
I have an IT and business background and have worked with databases in the past. Uganda needs to wake up.