Group 1 Sticky Notes

[arquicanedo]

Scoping

• Find use caes
• Engage SMEs to identify questions
• Use-case (scope). 2 tracks? 1) knowledge discovery, 2) infrastructure
• Data therapy with SMEs
• Find MVP that lets you design upper ontology without having to model the whole world
• Identify and profile datasets
• Key questions: Graph: x-axis (Effort), y-axis (Value)
• Upfill engravers (?) from relational thinking
• KG product teams: SME + ontologist, data engineers, presentation + visualization
• Keng (?) formalize SMEs input using DFL

Design

• Identify existing ontologies
• Top-level (upper) ontology design
• Find and capture unicorn (knowledge engineering?)
• Identify tool needs: taxonomy vs ontology vs data graph

Develop

• Define deliverable (& quality)
• How to K.I.S.S. on huge field of KG?
• Jobs-to-be-done framework
• Identify key "things" that allow to attach other sub-domains
• Do have good tooling support for KGs - ontology browsing, easy modification, etc.

Deliver

• MUP/POC for buy-in
• Scalability by design
• What does "done" mean? How do you define a successful deliverable for ontology design
• Interactive app to test hypotheses
• Make insights consumable
  • Notifications
  • Existing Apps
  • Easy-to-use visualizations

Review

• How do you test your modeling?
• Learn to iterate between scope and review
• Peer review vs final review
• Validation of the domain
• Post-release metrics and feedback
• Lifecycle management (Data assets & infrastructure)

[bolerio]

[bolerio]

[bolerio]

[bolerio]

[bolerio]

[bolerio]

[bolerio]

[bolerio]

[bolerio]