Talk:Data Quality Improvement Solution Offering

From MIKE2 Methodology

--Sean.mcclowry 06:59, 8 May 2007 (EDT) It may makes sense to merge this with the Data Investigation and Re-Engineering Solution Offering due to the high areas of overlap. This offering would use major elements of that offering and the Information Governance Solution Offering.

DQI Activities and Tasks Review notes

Below are the notes from an email exchange between Sunil Varanasy and Harshal Bhalerao. The decision on Sunil's suggestions is yet to be taken.

1. Data Use Categorization - Typically identify and categorize data usage across different lines of business, data use, subject areas. This helps to work with key system / data owners and identify important data. End result would be a data use matrix

HB: Activity 1.4 Organizational QuickScan for Information Development

2. Critical data elements identification and prioritization: The next phase would involve working with these owners as identified above to identify data elements for DQ review and prioritize these elements based on criticality

HB: Activity 3.5 Business Scope for Improved Data Governance

3. The next important step is to conduct a qualitative assessment which in effect conducts a survey to gauze or gather perspective on data quality for the data elements. Based on the results, we recommend areas for additional scrutiny and data profiling.

HB: Activity 3.6 Data Governance Metrics

4. The next step would be the actual profiling of the identified critical data elements. This would involve source system mapping, performing a data lineage, gathering source system access and conducting the profiling / investigation. Key profiling metrics such as DQ completeness, validity, consistency, accuracy etc should be compiled and analyzed.

HB: Activity 3.11 Data Profiling

5. The next logical step is to review the profiling results, data lineage and recommend a Trusted Source for each data element. Trusted Source is the source which gives the most comprehensive results based on step 4 above. The profiling results along with various system operational statistics such as data retention, data use, system accessibility should determine Trusted source and method of use ( operational reporting, BI reporting, analytics etc)

HB: Activity 3.12 Data Re-Engineering (The Trusted Source could possibly be added as a task)

6. Another important step could be to identify and create organization wise standard definitions for the above data elements which will in turn help the organization in its data governance, metadata management, master data management etc.

HB: Activity 3.12 Data Re-Engineering (Standardization task)

7. The whole DQ exercise should be wrapped up with a training curriculum and a repeatable methodology such that organization can conduct DQ investigation on a periodic basis using the existing infrastructure and resources (create a repeatable process).

HB: Activity 5.1, 5.11 – 5.15 (several activities in the Operate phase are included for user support procedures, training, continuous ongoing improvement)

Overall, I think this is a good initial draft but will need some tweaking so that we stitch it together as mentioned above and present a logical flow. From the activities mentioned above, can you pls. logically arrange them into design, development, testing etc.

HB: I agree. Once the additions are finalized, we can work on making the updates in the appropriate phases of the Usage Model on MIKE2.0 as well as the project plan.

Also, the project plan identifies some of the steps mentioned above at a broad level. for Ex. conduct profiling would have some predecessor and successor steps which are not explicitly stated in the plan. However it also mentions various other project management activities in great detail. So I guess we should balance it a bit here and synthesize the plan from current 1000 tasks to about 500. .

HB: The BE Delivery Framework project plan template requires us to keep the key project management activities and tasks. This does not mean the practitioners will have to use them on every project. This is to be treated as a suggested project plan and appropriate changes can be made by the project teams as found appropriate for the engagement.