Data Governance PMC Meeting 17th January 2019

Data Governance / Egeria Weekly Meeting - 17th January 2019

Attendees

• Mandy Chessell
• John Mertic
• Chris Replogle
• Ernie Ostic
• Maarten Steen
• David Radley
• Michael Trolier
• Suresh Channamallu
• Maryna Strelchuk
• Andrew Padilla
• Susan Malaika

Agenda

• Updates from the team
• Review of the Subject Area expert as an Egeria consumer
• AOB

Minutes and actions

• Mandy been working on architecture/maturity model for rolling out a governance program.
• John has been working on conformance and relaunch of marketing material.
• Chris working in intergrating Egeria's repository interface into the SAS runtimes.
• Ernie working with clients, IBM product teams and metadata bridge.
• Maarten wanting to learn more about Egeria as in integration technology for e
• Suresh looking are the infrastructure support Egeria
• Micheal looking at designing an enterprise metadata capability
• Andrew wanting to learn more about Egeria.
• Ideas from the review of subject area expert as a consumer (see below)
  • The subject area expect need support from senior leadership - can we help with justification.
  • How can the organization consume glossary terms to increase the perceived value.
  • How can the glossary be publicised.
  • Metrics are a key way to raise the awareness of the glossary.
  • Reporting queries on the quality and quantity of terms.
  • Which terms are in use in production.
  • How many assets associated with assets.
  • How would an information architect work with the subject area expert.
  • The author, owner, status of the term.
  • Compliance with ISO standards.

---

Review of the Subject Area expert as an Egeria consumer

In the Coco Pharmaceuticals company, a persona involved with creating Subject Area content is Faith Broker. As Jules (CDO) sets up the vision for the company around being data centric to get control of the governance, Faith takes on the role of of privacy officer. Initial tasks that Faith is involved in is to understand the business, which leads to a restructure of the way that data is organised into subject areas.

Other domain experts in Coco Pharmaceuticals include

Callie Quartile empathy map

Tessa Tube empathy map

Stew Faster empathy map

Faith Broker empathy map

Some thoughts on the next level of detail relating to their Subject Area role, using Faith as the example:

Think

• The systems that she deals with hold information with different names, difficult to know how data between systems relate
• Domain experts have a vocabulary that they use in their heads; Faith thinks this vocabulary should be written down semantically so it can be used the drive the business.

Feel

• Frustrated that there is a lot of important data kept in spreadsheets, with column names that are made up by the spread sheet author or copied from a technical source like a database column name. So it is difficult to know what the data means, and data can only be used for the purpose it was authored for. Faith hopes that finding common meanings allows data authored for one purpose can be used to create clear communication between departments and for other purposes; for example analytics - to gain insight into the business.
• The data from the systems she needs to deals is technical and complicated; Faith is frustrated that she need to map this into her view of the business.
• She is scared that the work on glossaries might be viewed as slide ware and be ignored if the organisation culture is not receptive.

Behaves

• Attention to detail at the conceptual level.
• Checklists or utilities to ensure that attributes and objects are well described
• communicates well the benefits of glossaries.
• organises workshops to create and maintain glossary content.
• ensures that the glossaries that are agreed upon are operationalized

Says

• Using the same language makes our business more efficient; leading to improved communication inside the company and to our customers.
• Data is most useful when you what it means.

Required skills :

• Good communications skills to influence the other members of the organisation to :
  • buy into the importance of having a glossary.
  • draws appropriate diagrams to show the glossary content
  • agrees the way that the glossary value can be measured.
  • Works with Jules Keeper to ensure that business is driving the data strategy not IT by:
    • working with all data stake holders to embrace the glossary to drive their data requirements where possible.
    • enhance discovery, consumption tools to be able to take glossary content.
  • organises ongoing workshops to brain storm on the terms that represent and should drive the business.
  • works with the team that maps glossary terms to assets to ensure there is not an impedance mismatch.
• Structural modeling (has-a, is-a-type-of relationships etc)
• Semantic modeling (Synonyms etc)
• Common information modelling
• Reference data
• Understanding the business with regard to what needs securing, in terms of retention, criticality and confidentiality.

How to acquire the required skills

Egeria github Data Governance Github Others?

What does Success look like?

• As with Jules Keeper, Faith wants glossary driven governance to effect every employee, so governance is unavoidable and makes work easier and business regulation is built in.
• Common information model used through out the organisation, reducing ambiguity in communications.
• metrics around glossary usage showing increased adoption and benefits - showing compliance to regulations or not. Ideally wants a dashboard.
• core business is completely (or to a large extent) driven off the glossary.
• leveraging reference data with the glossary content, so data like country codes can come from authorative sources and be used consistently in the organisation.
• vibrant community around data
• feedback loops & process in place to spot ill formed glossaries, so the glossaries can be corrected by a curator.
• glossary terms being used to drive processes in the organisation
• increased automation
• more and more parts of the business using tools that take and use glossary information.
• Tools / UIs in place to author glossary content
• Tools / UIs in place to find, explore then bring in glossary content - embedded in existing tools.

What does Faith do to boot strap the new data strategy?

• create the subject areas - link
• brain storm the initial important terms
• identify a short term project that can show the benefit of the new approach over the existing approach
• run workshops to flesh out the structural glossary, and ontologies, then authoring this glossary content in an appropraite tools.
• set up appropriate workflows to ensure that glossary creation, updates deletions are appropriately reviewed prior to being put into production.

Semantic organisation maturity

Initially an organisation strives to use common terms to drive the day to day business After this it strives to surface the terms in many tools by making terms findable Semantic information is used to drive information governance including access control.

How mature is the Egeria Subject Area OMAS.

Basic authoring use case

1. simple (Create,read, update,replace,delete, purge,restore ) authoring of Terms, Categories and Glossaries - Done
2. simple (Create,read, update,replace,delete, purge,restore ) authoring of term relationships - Done
3. Get term relationships - Done

Basic authoring use case and finding where to start

4. Find Term

• Other finds
• Other get relationships

Basic authoring use case and finding where to start

• collaborative API (including projects comments, TODOs likes)

More sophisticated use cases

• API exposed around spine objects
• API exposed around governance ownership management

Graph like API

• API exposing nodes and lines so resource can be easily manipulated as a single type in a graph.

Events

• events - partially coded - not tested

Misc

• Define default Icons
• remote glossaries
• projects

UI :

• defining subject areas
• brain storming glossary terms
• creating taxonomies (category hierarchies)
• creating ontologies
• glossary exploration and navigation

UI notes:

• forms could be used to create Term properties and the like.
• spreadsheet style of browser interface is quick and easy text based way to create terms
• tree and canvas, drag and drop style is useful. 2 Styles are useful:
  • navigation and exploring: e.g. driven by a query like show me all the content 3 deep from a term
  • authoring: intuitively wire together relaitonships Note dashboard and monitoring graph interfaces are likely to be of use as well, less so for the subject area expert.

Outreach

How do we bring Faith, Tessa, Stu and Callie into the ODPi Egeria community?

• Given this audience:

  • Names of jobs that would author subject area? Chief Privacy Officer, Domain experts, and?
  • maybe roles are not yet present so need to be created by the CDO
  • associated groups:
    • academics
    • standards bodies around semantics?
    • governance conferences for CIO and CDOs.

• Where are that audience?

  • Universities
  • Which conferences?
  • Egeria open source community

• How do we reach that audience?

  • Conferences
  • marketing - which channels are effective.
  • social media
  • do we want to target specific industries?

---

License: CC BY 4.0, Copyright Contributors to the ODPi Egeria project.