Analytic Tools for Data Warehousing!
Now that you got the data warehouse schema set up! Now for the to(ols)ys!
There are many kinds of analytic tools that you can use in your organization with your data warehouse, so instead of listing them out, I’m just going to discuss the overview on certain tools on why they are useful and how they can help. In the architecture design post, I mentioned that there are various front-end tools used with the data marts so that is what we’ll be looking more at today. As a result, these tools will be used in forming strategic and tactical business decisions by ensuring that the right information goes to the right person through the right channel at the right time.
As you can see, there are a variety of tools that are used in this schema, primarily in healthcare. In some cases, the administrator can set up are automated alerts. These alerts can then be used to notify when certain thresholds and results are reached or on the other hand, if a certain result is not attained. The rules can often be adjusted as needed. One such example can be with EHR, and that alerts can be sent every time a record is not fully qualified to the highest level (useful in documentation and reimbursement purposes), that way these entries can be flagged for administrators to go back into and see what was lacking or wrong. In doing so, it allows a defense layer to identify business issues quickly and to provide notification in order to possibly avoid serious problems.
Another analytical tool that can be used, are data mining tools. I didn’t want to name anything specifically, but data mining tools are useful to help discover underlying correlations or to analyze more obscure pieces of data to help gain business intelligence by identifying and observing trends, problems and anomalies. Data mining tools are great because they can help to quickly identify emerging patterns and trends. I know for me, the majority of the times I’ve seen the word “data mining” has been when people analyze the patch notes or code of new game updates to find out more about new items being added and content haha. But in the real world! Data mining can be used to help find trends and correlations, similar to statistical analysis in regression curves. Although it can be easy to find that people who purchase diapers often purchase beer as well, it can be taken one step farther and find that young husbands are frequently asked by the spouses to pick up diapers and other household items as well!
Another one of the analytic tools isn’t exactly too unique, excel spreadsheets. Spreadsheets are a powerful, flexible and inexpensive tool that can help with the presentation of data from data marts. Even before data warehousing, spreadsheets were popular because they were able to organize data sources, albeit manually at the time. However, with these spreadsheets they could be copied and taken home and thus exposed security vulnerabilities and concerns. However, today, these spreadsheets could “mail merge” with the data warehouses to automatically populate and update these spreadsheets that can then use the data for graphing and other presentation methods. Another method is OLAP or OnLine Analytical Processing which is a tool to analyze multi-dimensional data to help generate reports as well. OLAP has seen a rise in popularity because of its ability to replace static, paper-based legacy reports with online access corporate information via OLAP!
Web Analytics, help by focusing on the interaction of users with an organization’s website and content. It leverages data to obtain business intelligence and can thus allow the organization to analyze user data from web server logs to integrate with the data warehouse to allow an even greater leverage. This one of the analytic tools allows BI to identify technical and navigation issues, better understand user’s unique needs, interests and patterns, and lastly, identify improvements for website design.
When a request is initiated by the web browser, identifying data is sent with the request. This data allows the server to return the web page, however the server logs all of the information into its data warehouse. And because of the nature of the data it can be sliced and diced as needed to generate BI. Additionally, there are often cookies and session IDs that can be used. By doing that you are able to track the IPs, user counts, visits, time stamps etc. Even better when an organization has an authentication capability, because then even more information is able to be logged to help generate BI. By having this kind of information, it can help cater changes for repeat and potential users/customers. Aside from the packet contents, another example that can be used to track web access is a dimensional model using data consolidated from a web server log, an authentication application and a database.