Michael Briles and Joanne Dority, SAP provided an ASUG webcast recently on Metadata Management.
Agenda:
- Challenges with data quality and integration points – what are pain points
- Product vision and value proposition of Information Steward 4.0
- Key Capabilities
- Demo
Information Steward 4.0 is in ramp-up.
Challenges with data quality and integration points – what are pain points
According to Michael, there is no environment for business users to collaborate with IT regarding data issues. There are times when IT will understand there is a problem with the data, and will address this with the business – there is not a common way to collaborate. The business user wants to know where data on a report comes from, too. Users are not sure what the business definition is - for example at one company they could not agree on definition of what a store was.
Product vision and value proposition of Information Steward 4.0
Figure 1 Vision – Source: SAP
Figure 1 shows the product vision for Information Steward. The first part of the product vision is to “empower data stewards with a single environment to discover, assess, define, monitor and improve the quality of their enterprise data.”
Information Steward is defined into three topic areas. One of those areas is “Discover” (discover and understand enterprise data) comprised of data profiling and metadata management and ability to catalog the data assets.
The other area is the “Define” or define data ownership, assign data ownership, accountability and roles. Here you define business terms, validation rules, cleansing rules, and models. You might have customer data in multiple systems; are people signed on to do data governance in those systems? Do they own the data and are responsible for owning the quality of the data?
The last area is “Monitor and Remediate”. Once we have discovered the data and defined who owns the data, then you want the ability to monitor the data quality. The result of this monitoring is to surface data quality score in business user applications. Finally, you need workflows to resolve data quality issues within Information Steward Application.
Most (but not all) is covered in the Information Steward 4.0 release.
Discover area is covered with data profiling and metadata management.
Define area allows you to define business rules.
In the last area you can monitor data quality and surface quality scores. There is still work to be done with workflows to resolve data quality issues.
Figure 2 Why Information Steward – Source: SAP
Figure 2 shows three primary areas. How good is my data and Where did my data come from?
For Data Governance, NetWeaver MDM is a roadmap item.
Transparency – if data quality is poor, business users want to drill down from a dashboard and see it and why is it bad, including answering if it is bad all across systems.
The vision is one integrated metadata management, business glossary, data quality assessment, data quality monitoring and cleansing package builder solution. This is what is being delivered today with the Information Steward 4.0. This is one place for the data stewards and business analysts to collaborate and govern data assets.
Key Capabilities of Information Steward 4.0
Figure 3 : Source: SAP
Information Steward Is in Ramp-up and generally available after ramp-up.Information Steward is a new product but also a successor product .
Figure 3 shows the tabs. Data Insight has new functionality added to Metadata Management. Have an existing 3.x Data Insight product. Purpose of the product stays the same – ability to data quality monitoring and data profiling. It is a brand new product
Metadata Management is largely unchanged functionality from 3.x. This would be an upgrade for existing customers.
Cleansing Package Builder is part of Data Services / Data Quality Management. Rather than be in Data Services, CPB is now inside Information Steward 4.0. This is migration from one application to another.
Metapedia is a business glossary and comes from Metadata Management and will migrate to Information Steward 4.0 release.
Data Insight
Figure 4 Data Insight – Source: SAP
Figure 4 on the left shows the scorecards for four data domains. It shows the data quality score. Thresholds are configured by the data steward. Different quality dimensions are part of this score, such as accuracy, completeness, conformity (how well does the data conform), uniqueness, and integrity (is address valid)? The line graph below is the quality trend and show pre- and post-data governance scores. Quality trend measures quality over time.
With Information Steward you can drill down into that information to understand where the poor quality data is and see the records. 
Figure 5: Source: SAP
Figure 5 shows a scenario where you are already using NetWeaver Master Data Managment (MDM) and point BusinessObjects Information Steward towards MDM. This is not out-of-the box functionality. There are no dependencies between the different applications. BusinessObjects Information Steward is a completely stand-alone product and does not require MDM or Data Services.
Metadata Management
Figure 6: Source: SAP
Use case is “where does the data come from?” – This is your data lineage. A user accesses a report in BI and the user does not understand where the data comes from.
You can do the reverse which is the impact analysis. How will a change in the source impact the reports and downstream.
Cleansing Package Builder
Cleansing Package Builder (CPB) is functionality migrated over from Data Services and surfacing inside Information Steward. Michael said that CPB empowers the data steward to drive their data cleansing solution rather than go into Data Services.
An example would be to develop data cleansing solutions for data domains that SAP does not provide out-of-the box such as product data in manufacturing.
Figure 7 – Why Need Cleansing Package Builder – Source: SAP
On Figure 7, on the left we have a data source that comes in as a single text string and you define how that information would be parsed out and characterized on output on the right.
Metapedia
Figure 8 – Metapedia Use Cases – Source: SAP
Figure 8 shows the key drivers for Metapedia. How do we define a store for Starbucks? How do we establish ownership for key data elements? This work is done by the data steward in Metapedia.
DEMO
Joanne provided a demonstration of Information Steward and an overview of its touch points into BI.
Figure 9 Information Steward – Source: SAP
Figure 10 Metadata Management – Source: SAP
Figure 10 shows all metadata that has been collected for this Information Steward instance. Each box represents a type of metadata.
The first box on the upper left is from the BI area from the BusinessObjects Enterprise which is a Rapid Mart. Reporting data has been collected.
The box in the upper right is the Data Integration which contains the ETL tools, Data Services and third party ETL.
Relational databases and modeling tools – there are over 50 different sources of data to collect from. The directory view lets you browse and maintain the metadata.
Figure 11 – View Impact – Source: SAP
Joanne showed some of the touch points of metadata management when she went to the Data Insight dashboard shown in Figure 9, clicked into the accuracy quality dimension (which isn’t doing so well). Figure 10 shows where you can view the impact of this poor score. If this is a poor data source, who is viewing and using the data source.
Figure 12 - Source: SAP
Each box in Figure 12 shows different steps of where the data is consumed. Left is the starting point and to the right is who is consuming the data.
Figure 13 – Source: SAP
Figure 13 shows the data flows from Metadata Management
From Figure 13 the flow is from Data Services to BOE in Figure 14, which is a Universe, used in a number of reports. If the quality score is low you can see the overall environment.
Another touch point is the report itself.
Figure 15 – Source: SAP
If you are looking at a report and wondering where it came from, how are the numbers calculated and what source system did it come from?
Figure 16 – Lineage: Source: SAP
These diagrams in the reverse, showing all the steps the data went through prior to being placed on the report.
Figure 17 – Source: SAP
Figure 17 shows how net sales was calculated.
You can create your own attributes to add to the existing metadata. You can also edit annotations.
Figure 18 Metapedia – Source: SAP
Metapedia has the ability to capture business terms. Figure 18 shows categories of business terms on the left. You can import via Excel if you want.
Figure 19 – Source: SAP
Figure 19 shows the definition of VAT. You can relate terms to other or associate to technical metadata
--------------------------------------
Question & Answer
Q: How much is visible to casual user – can they see these descriptions?
A: You can add a Metapedia tab to the report
Q: Can meta data management go to the ECC BW extractors, or better yet back to the source ECC tables under the ECC BW Extractors
A: Can go to BW, but the ECC side is a roadmap item.
Q: What is the export/integration functionality from Metapedia? Thanks!
A: You can export to Excel. You can place your own SQL queries against the repository.
Q: Does BOB-IS has out-of-box connectivity to SAP NW MDM 7.1?
A: No, not out-of-the box. It would be an implementation/configuration scenario.
Q: Is Information Steward restricted to Windows only as BOMM is?
A: Primarily Windows-based
Q: Can you extract the metadata information from Microsoft SSIS ?A: Yes
Q: Do you have to be a BOE 4.0 ramp-up customer to do an Info Steward ramp-up customer?
A: No
Information Steward is in ramp-up and there is a ramp-up nomination process and contact your Account Executive.
Special thanks to the speakers for a great webcast and Ina Mutschelknaus, SAP, for arranging such great webcasts for ASUG. Follow Ina on twitter at @InaSAP.
- SAP Managed Tags:
5 Comments
