Let's explain this in more detail: The AX web client (shown in the bottom right of the architecture graphic as "web browser") offers browser-based access to data analysis projects which are stored centrally on the company's (own) server. These projects are collections of elements, e.g.
- Raw data
- Analysis routines
- Related files
Before looking at the 6 options for the web client in more detail, let's briefly explain these four terms:
These data are extracted from a source system (e.g. SAP® ERP ECC 6.0) and are available as a copy on the analysis server. The data could, for example, be data relating to a supplier base and accounts payable, and also purchasing documents. In the architecture graphic above, these data were extracted from the source system on the left and are available on the central analysis server.
These are predefined analyses which are started either manually or automatically and which analyse the raw data and generate results. Examples here could include a duplicate payment analysis, which is automatically performed each month based on accounts payable data and the result of which is produced as a list of potential duplicate payments. In the graphic above, these are located on the central server together with the raw data.
By filtering or compressing generated data, ideally suited to give an overview of a dataset or to understand situations at individual document level. If for example the aim of the duplicate payment analysis is to give the accounting department a list of potential multiple payments for processing, then the individual transaction becomes a focus. The aforementioned flexibility also comes into play: The results can be provided in different ways, ideally tailored to the user's preferred method of working. Tech-savvy users can use the ACL™ desktop analysis tool and Excel™ specialists their beloved Microsoft Excel™, while browser-based access undoubtedly presents the easiest option. These three options are indicated on the right-hand side of the graphic.
These can be data which are automatically directly included in the analysis logic (e.g. a file containing internal suppliers, which are not to be considered during the duplicate payment analysis) or additional information such as analysis documentation or similar accompanying material. I have mentioned these for the sake of completeness – they would be stored centrally on the server and every user can view/download them.
Data analysis of the huge source data volumes is thus performed centrally on a server, ideally as automated as possible. Results are then generated in different ways, depending on the concept and on user requirements. In this blog post we are focussing on the possibility of browser-based working with data and on simple but effective visualisation. Strictly speaking, this goes beyond simply generating graphics, but we will explain this in more detail later.
Possibilities for analysing data via a web browser
Moving on from this brief diversion about the architecture and the technical possibilities offered by central data analysis based on a client/server structure, we will now look at the opportunities offered for working with data in a browser and for improving the ability to interpret results:
- Viewing data
- Filtering and sorting
- Conditional formatting
- Generation of graphics
- Adjusting the view
- Re-usability of the interpretations
The first five points are options which can be used for the data on an ad-hoc basis. In point 6 I will then go on to explain the concept of "interpretations" in this connection, which makes it possible to re-use points 1 to 5.
As an example, I will use the 11,268 order items from the introductory Excel example. This will serve as an analysis result, for which we want to obtain an overview.
The data records can now be viewed. In the case of larger volumes of data, the table is not loaded in its entirety but is loaded gradually as the user scrolls the data.
In contrast to purely locally installed analysis solutions, a client/server architecture enables data and results to be made available to users in the most suitable form for them. A fresh approach here is browser-based access to data and analysis results. This offers simple but constructive ways of interpreting results using metrics, graphics, conditional formatting, comfortable filtering or adjustment of views. The sum of these possibilities can be stored and re-used as interpretations. The aim here is not to replace special tools such as the Tableau visualisation software, but to support the data analysis process as much as possible – from source data to analyses of the detailed results through to interpretations. Although a user may be faced with huge data volumes, this makes it possible for him to see the wood despite the many trees – in other words, to draw the right conclusions from his data and identify interesting facts without having to worry too deeply about technical aspects.