Using Knowledge Graphs to Document Cloud Native Products
- Partner presentation
- Intelligent Delivery of User Information
- 05. November
- 11:30 - 12:15 AM (CET)
- C6.1
- finished Presentation Slides
-
Martin Blumbach
- Ericsson
-
Stefan Reinhardt
- eccenca
Contents
Cloud native software products consist of individually life-cycled, frequently released, and flexibly assembled components called "microservices". This presentation explores how technical documentation for such products can be automated. The challenge for automation is that content must be life-cycled per microservice but must be presented from the perspective of the product user who "sees" functions and use cases, not microservices.
We'll present our content delivery pipeline built around a "knowledge graph platform" based on eccenca corporate memory. Knowledge is expressed in RDF knowledge graphs automatically instantiated from machine readable information "fragments" and manually created DITA topics. For publishing, we automatically assemble DITA bookmaps including manually and automatically created DITA topics. We'll also present visualization of knowledge for SMEs and give an outlook on how to repurpose knowledge, for example for Generative AI.
Takeaways
Using knowledge graphs in technical documentation:
- is a marathon, not a sprint
- requires active change management and competence buildup
- allows automation by making tacit knowledge explicit
- opens up new opportunities
Prior knowledge
- DITA basics
- Knowledge graphs (not necessary, but beneficial)