November 12, 2010
However, it is growing much faster that our ability to analyze it. And still the analysis of the data (often inconsistent and messy) is a crucial in every field. Today there are many tools available to help analysts organize their data, and in many cases it greatly helps, however it cannot meet the demands of our age – the analysts just cannot react properly to the constantly changed situation.
It is here where interactive analysis stands out as the mean to better and faster derive insight to the data and thus make better and timely decisions.
The basic idea of data analysis is to represent information in a human-friendly way, highlighting useful information, suggesting conclusions and ultimately supporting decision making. Data analysis assumes many techniques that can be used to analyze business information: OLAP, data mining, business intelligence, visual analysis, etc. The final part of any data analysis technique is producing reports answering questions of interest. There are many report generating systems that help people get their data in most easy way, but many of them only represent information in a one-time slice form, and so the analysts may only see what they are given, possibly missing all-important recent info.
And here comes the idea of interactive reporting.
The key to the success in decision making is in having right and complete information when you need it. And here is the list of “must have” requirements for a report generator to make this possible. First, it must be able to take data from different data sources and combine them in a single report. So even if the data is located in different places the user must get them all in one place, combined and prepared as the author defined.
Then it should always provide “fresh” data. It means the report must be always actual and contain up-to-date information. Practically it means that the author should be able to define the way how and the time when the data in the report is refreshed. The end user simply does not have to know about it. What he needs is always having fresh data at hand.
Next, the report should be able to contain different elements or controls (widgets) combined in any order and possibly connected to different data sources. These widgets can display different or even the same data in different ways. Say, we want the report to display the same sales data as a sorted table and below have the diagram showing how these sales has changed over the given period of time.
Also there must be “active” or “live” widgets in the report. OLAP Grid can be a perfect example of the like, because it allows the user working within the control: choose the active data slice, “drill down” to more details, “drill through” the selected cell, and more. So it provides interactivity to the whole report and ability to select the data in view without generating another report.
And finally the report should allow different extents of freedom to different users. Depending on the viewer the report should give different levels of access to the data, different data sources, different widgets available, and different interactivity levels.
As a result the users get the full range of info they are really interested in, and they can actively use it to analyze in many possible aspects. If allowed, analysts can access more information as they go. The same report can be used by many users with different access levels. The levels can be controlled by the author of the report to address different audience. All this takes the analysis process to the next level of productivity and ultimately helps decision making.
John Parsons is on the staff of Radar-Soft LLC. Founded in 2004, Radar-Soft is developing OLAP and Business Intelligence solutions for developers in different platforms (ASP.NET, NET Windows Forms, WPF, Silverlight) and analysts. Now Radar-Soft also offers a full-featured BI-reporting solution for end users called RadarBI. You can know more about the Radar-Soft at www.radar-soft.com.