Building Out The Ontology Pipeline

Jessica Talisman's excellent article, The Ontology Pipeline, makes two very important points that I see.

First, I have historically and incorrectly seen what Talisman more precisely sees as a dynamic "pipeline" as more of a static "spectrum".  For example, the first two graphics on this page I had always looked at the graphic as an "either/or" choice that had been made once.  You either build a "controlled vocabulary" or a "thesaurus" or a "taxonomy" or an "ontology" or a "theory".

But what Talisman seems to be saying is that this is not a one time choice, this is more of an "evolution" from one point to another point within that spectrum tools.  The tools are not "alternatives", they are "steps".

The second important thing that Talisman is pointing out is that the domain of library science which has been working with these sorts of things for 10+ years offers proven, repeatable methodologies for constructing scalable, resilient, extensible, accurate, and therefore reliable knowledge management systems which can then be used effectively by AI systems.

National Information Standards Organization (NISO) publishes the standard, ANSI/NISO Z39.19-2005 (R2010) Guidelines for the Construction, Format, and Management of Monolingual Controlled Vocabularies, which are such guidelines and conventions.

Using these proven methodologies which are battle-tested by library science to document your data which turns that data into information which then can be used by things like axiomatic systems that work like a Swiss watch.


Additional Information:

Comments

Popular posts from this blog

Big Idea for 2025: Semantic Accounting and Audit Working Papers

Microsoft CEO: "AI Agents will Replace All Software"

Professional Study Group for Digital Financial Reporting