SAP BW Data Mining Analytics: Process Reporting (Part 2)

Summary

SAP BW Data Mining serves as a course of design platform for all types of analyses each based on the  info mining methods (each geared up by SAP as built-in with SAP BW Data Mining or geared up by licensed distributors), or based on “non-scientific” transformations (joins and unions of data tables, sorting and filtering of data, grouping and aggregation, and so forth.). The interface of the Analysis Process Designer (APD) efficiency of  SAP BW Data Mining provides steady reporting and visualizations. Still, there might very effectively be a necessity for further process-related analytics which may facilitate administration and deployment of the content material materials created with SAP BW Data Mining. In this paper we’ll present the following analytics:

• Dashboard – SAP BW Data Mining Process Reporting

Business Requirements

The important use of SAP BW Data Mining is creation of fashions based on analysis methods and of examine processes based on the fashions (or, inside the cases of “non-scientific” analyses, not involving info mining fashions). Together, fashions and analysis processes are the content material materials that is constructed using SAP BW Data Mining. If SAP BW Data Mining is used actively in a company, the rising amount of the content material materials that is constructed might end in a necessity to have additional means for reporting on that content material materials.

The transaction RSANWB, which allows creating and modifying the analysis processes, includes basic efficiency for course of search (“Find” and “Find Next” buttons):

Although in numerous situations related to working with analysis processes this very seen efficiency might be adequate, we’d provide the subsequent additional enterprise requirements:

• Ability to browse explicit analysis processes to visualise their most important particulars with out clicking into these processes and discovering out them by the use of a Display/Edit mode
• Ability to visualise the processes that embrace a specific object (info mining model, DSO, and so forth.)
• Ability to visualise the processes of a specific mannequin (A, M or D)
• Ability to visualise the processes with a specific kind of course of/object affiliation (requires, sends info to, receives info from, and so forth.)
• Ability to visualise the fashions with objects of a specific kind (CUBE, DMMO, ELEM, IOBJ, ISET, ODSO, and so forth.)

Analytics

The “SAP BW Data Mining Process Reporting” dashboard permits to implement the above enterprise requirements. The BEx query that is supporting the functioning of the dashboard takes the information extracted from the SAP BW Data Mining’s tables that retailer the main points about analysis processes.

At startup, the dashboard reveals a listing of processes presently obtainable in our explicit SAP BW Data Mining event (see beneath screenshot). The following columns have been enabled in that file:

• Process ID – a novel identifier, the “technical title” of the analysis course of
• Process Object Name – an identifier of an object of the analysis course of
• Process Version – the mannequin of the analysis course of
• Process Association Type – the type of the affiliation between the analysis course of and course of objects (see beneath a fraction of the strategy affiliation kind value file)

• Process Object Type – the type of the strategy object in step with the BW object repository (see beneath a fraction of the strategy object kind value file)

The beneath screenshot reveals the file of examine processes displayed on the dashboard’s startup:

The selectors of the dashboard match the columns of the file and allow limiting the processes visualized by the use of the file to explicit requirements. Each time a specific value is chosen, the respective selector’s standing indicator turns inexperienced.

In the beneath screenshot, the PIO_MRO_PREDICTION_CL value is chosen by the use of the Process ID selector. The file of processes is immediately refreshed to point out solely the information akin to the analysis course of with the technical title PIO_MRO_PREDICTION_CL:

If we would have liked to accumulate the file of processes that embrace a specific course of object, we’d use the Process Object Name selector. In the beneath screenshot, the PIO_MRO_CL value is chosen inside the Process Object Name selector, and the file returns solely the processes that embrace the article PIO_MRO_CL:

In order to further refine the selection, we’ll maintain the PIO_MRO_CL object chosen inside the Process Object Name selector, and might proceed with making use of additional requirements. In the beneath screenshot, it is confirmed that inside the Process Version selector the value “A” is chosen, the following recordset displaying solely the analysis processes containing the PIO_MRO_CL object and having mannequin A:

To proceed with the refining of the following recordset, we’re ready to decide on a specific course of affiliation kind inside the Process Association Type selector. See the strategy affiliation kind 4 chosen inside the beneath screenshot:

As a outcomes of our refining, now we’ve ended up with only one file that meets all of the above requirements. We will clear the picks made sooner than inside the Process Object Name, Process Version and Process Association Type selectors – and might present the alternative picks centered on the strategy object kind. For occasion, we might wish to current the version-A analyses processes that embrace objects of DMMO kind associated to them by the use of the affiliation kind 4. In order to do this, we’ll take away the selection made inside the Process Object Name selector (PIO_MRO_CL) and apply the Process Object Type selector (see beneath):

Typical Use Cases

The following might very effectively be examples of the usual use cases by way of which the utilization of the SAP BW Data Mining Process Reporting dashboard would possibly convey benefits:

1) An group maintains a diversified library of examine processes, and it faces the necessity (e.g., on account of availability of additional superior prediction methods and fashions) to interchange the model A in all the analysis processes that embrace it with the model B. It is assumed that the model A could also be contained by analysis processes containing a extremely broad file of various course of objects (i.e. there is not an everyday affiliation with one or a set of various course of objects).

Use state of affairs: select the information that correspond to the analysis course of containing the model A using the Process Object Name selector.

2) An info mining specialist has carried out changes inside the buildings of all of the DSO objects associated to the current analysis processes, and now desires to manage accordingly the sector mappings among the many many DSO objects and all the fashions which might be sending info into them.

Use state of affairs: select the information that correspond to the ODSO object kind by using the Process Object Type selector, after which choose the strategy affiliation kind 4 inside the Process Association Type selector.

3) An info mining specialist wish to uncover out the strategy objects from which processes get hold of info.

Use state of affairs: select the information with the affiliation kind 5 using the Process Association Type selector, after which open the dropdown file inside the Process Object Name selector – the dropdown file might be exhibiting the strategy objects which might be involved inside the analysis processes as suppliers of data.