I’ve been thinking about a financial OLAP system based on XBRL/XML/Json documents (free and open source) , in which components based on Front-End advanced technologies are implemented.
The state of the art of this kind of system (free and open source) is based on MDX queries (ex.: Forensic LMDQL, [if you know another one, please let me know]), in which the financial analyst or auditor must know about this computer language (MDX), as following:
1 SELECT 2 { [Measures].[Store Sales] } ON COLUMNS, 3 { [Date].[2002], [Date].[2003] } ON ROWS 4 FROM Sales 5 WHERE ( [Store].[USA].[CA] )
The system views are based on Mondrian JPivot (a JSP-based OLAP Client), in which the user have to insert MDX queries to initialize the forensic analysis, as following:
So, this OLAP system has its processing based on statistical calculations for fraud detection over XBRL documents. All of those calculation are applied on non-computerized audit, henre, its feasibility is indisputable.
However, we can do a new approach about how the financial analyst or auditor can handle the XBRL documents through this tool. I mean, we have to figure out a user-friendly solution for XBRL data visualization for OLAP systems.
I have been created a prototype in which it’s possible to get the idea in which I try to improve the usability of both technology: LMDQL (and its financial OLAP operators) and Forensic LMDQL (and its financial forencic OLAP operators). This idea has used JavaScript technologies (as such: drag&drop), in which XBRL documents (and its data) can be a draggable object on web platform, as following figures:
This is the analytical module of the XBRLFramework System.
Thoughts? If you have interest about this project, please contact me.