Collaborative AI: Fast Track to Concept Maps – Turn a Flat Set of Concepts into a Knowledge Graph

  • Presentation
  • Terminology
  • 11. November
  • 14:30 - 15:15 PM (CET)
  • Channel C
  • finished
  • Presentation Slides
  •  Michael Wetzel

    Michael Wetzel

    • Coreon GmbH

Contents

Collaborative AI aims at developing systems where humans and artificial systems work together. Can this method help terminologists to turn existing yet still flat terminology lists into a structured concept system?

In a research study with EU IATE we compared a pure manual approach against semi-automatic concept map creation. We prove that the efficiency increase makes the taxonomization of even large terminology databases feasible.

Takeaways

How can Collaborative AI help for the speedy creation of concept maps? What AI / machine learning algorithms help to draft maps? What are the strengths and weaknesses in time, approach, quality?

Prior knowledge

Begriffsorientierte Terminologiemethodik; Wert und Nutzen von Begriffssystemen / Wissensgraphen

Speaker

 Michael Wetzel

Michael Wetzel

  • Coreon GmbH
Biography

Managing Director of Coreon GmbH, with a focus on merging concept-oriented terminology management, taxonomy and ontology products. Before, Michael was product managing TRADOS / SDL MultiTerm. His background is in Computational Linguistics. Active contributor in ISO TC37 to drive TBX forward.