The other month, Michael Briles, SAP, provided ASUG a webcast on the new Information Steward EIM 4.0 solution.
Agenda:
Challenges with Data Quality and Integration Projects
Why design a solution around Information Steward?
Figure 1, Source: SAP
Figure 1 shows the product vision. At the center is the Information Steward, designed into 3 different categories:
1) Discover – catalog data assets through data profiling and metadata management
2) Define – business terms, cleansing rules, assign ownership. If customer information resides in CRM, ECC, outside – who owns?
3) Monitor and Remediate – monitor data quality over time
Figure 2, Why Information Steward? Source, SAP
Figure 2 shows 3 driving factors. For Business and IT collaboration – what is the lineage for the report?
For data governance – where is our data quality at, and assign ownership, define corporate standards through a corporate glossary, do remediation within Master Data Management
For transparency, be able to drill down from scorecards. Michael said not just corrective, but also be able to understand root-cause.
Figure 3 –Source, SAP
Figure 3 displays what is delivered in Information Steward, 4.0, which is in ramp-up
Key Capabilities of Information Steward 4.0
Figure 4, Source: SAP
Four primary areas are delivered as part of Information Steward 4.0 as shown in Figure 4.
Data Insight is new functionality. Data Insight 3.1 was rewritten for Information Steward.
Metapedia and Metadata Management come from existing 3.1 solutions but moved into Information Steward.
Cleansing Package Builder was part of Data Quality package.
Figure 5, Use Cases for Data Insight of Information Steward 4.0, Source: SAP
In reviewing Figure 5, Michael said it is not just about correcting the data but validating the processes, procedures and controls you have put in place. Figure 5 shows the scorecard information; we have 4 data domains. Within each of these scorecards we have the data quality score covering integrity, accuracy (does zip code match city?). Below you can see the data quality trend to see how it is doing over time.
Figure 6, Source: SAP
Figure 6 is a use case. It is not delivered out the of box. It shows 3 components working together: Information Steward, MDM, and Data Services.
Figure 7, Source: SAP
As Figure 7 shows, the functionality of Metadata Management remains unchanged; it is simply being moved to Information Steward. The key question it answers is “where did my data come from?”
The reverse of data lineage is impact analysis – how will this change impact the reports?
Demo
Abhiram Gujjewar, SAP, then provided a demonstration of the solution.
Figure 8, Source: SAP
Figure 8 shows the entry point to Information Steward is just like the slides that Michael showed with the 4 tabs. Abhiram then clicked into the scorecard of the Sales Operation Project.
Figure 9, Source: SAP
Figure 9 shows the scorecards for the Customer and Product data domains of the Sales Operation Project. The overall Quality customer domain score is 7.88 but notice that Conformity is not doing well with a score of 5.95.
The Product overall score is 9.16 as Completeness and Integrity scored at 10 each.
Abhiram then clicked show more to see the details behind the scores.
Figure 10 Details of Data Insight, Source: SAP
Figure 10 shows the details behind the data quality score, including the validation rules and the rules. We can see from Figure 10 that there are data issues with the e-mail address and the phone number.
Figure 11 Drilling into Accuracy Dimension, Source: SAP
Abhiram drilled into the Accuracy dimension in Figure 11 to show that there are problems with the Country/Region field.
Figure 12 – Deeper Dive into Telephone Number issues - Source: SAP
Abhiram drilled into the customer telephone number conform to standard validation rule and it shows a graph of the issues, and how those issues are trending over time.
Figure 13 - Source: SAP
Abhiram drilled into the data issues itself as displayed in Figure 13.
Figure 14 - Source: SAP
What is the impact of this bad phone data? Figure 14 shows where it is used and what applications. This helps you determine the impact of a poor quality score (which reports and applications use this data element?).
Figure 15 - Source: SAP
Figure 15 shows the rules and the statistics behind what failed.
Figure 16 - Source: SAP
Figure 16 shows how to set up the scorecard.
Subset of Question & Answer:
Q: Does that mean Metadata management is no longer a standalone product and merged with IS or it's still a standalone product with new versions would be coming in future?
A: Metadata management will follow its regular maintenance cycles in terms of service pack and fix packs, but there will not be any Metadata management 4.0. It is essentially Information Steward 4.0
Q: What is your description of the "data steward" role. Is this an IT or Business role?
A: Data steward leans more towards business than IT. They know the data domain.
Q: How are CP (cleansing packages) deployed to DS?
A: There is a specific Publishing process in CPB (cleansing package builder) to "deploy" it to DS (Data Services)
Q: To use Info Steward 4.0 do you need BIP (aka BOE) 4.0 platform separate license?
A: Business Information Platform is included in the licensing of Information Steward 4.0. You will also need Data Services 4.0 in order to execute packages defined in Information Steward 4.0 Cleansing Package Builder. When you install IS 4.0, the installation process installs mini versions of BOE and Data Services.
Q: Can Data Insight test rules against ERP directly?
A: Yes, Information Steward 4.0 can test rules against ECC directly.
For related information, see Ina Mutschlknaus's blog this week about VA Hospital Spurs Change Through Data Transparency.
Thank you Michael Briles and Abhiram Gujjewar of SAP. Special thanks to Ina Mutschelknaus, SAP, for reviewing these notes and setting up such great webcasts for ASUG.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
11 | |
9 | |
7 | |
6 | |
4 | |
4 | |
3 | |
3 | |
3 | |
3 |