Generative AI for Taxonomy Creation (with abstract)

Article published by [[Other sources |Enterprise Knowledge]] and reviewed by Nikolay Sidelnikov
Generative AI for Taxonomy Creation (with abstract)
Published By[[Enterprise Knowledge]]
Publish DateApr 8, 2024
AuthorsHeather Hedden
Topics AI
Article Sizewords
Downloadable VersionNo
Page OwnerNikolay Sidelnikov
Summary AuthorNikolay Sidelnikov


The article Generative AI for Taxonomy Creation by taxonomy expert Heather Hedden, published by Enterprise Knowledge, contains an analysis of GhatGPT applicability for taxonomy creation. The author identifies the strengths and weaknesses of generative AI in the taxonomy creation context and points out where these tools might be utilized and what risks they cause. The article would be useful for experts who are responsible for digital commerce catalog design and CX designers.

First, the author decomposes the process of taxonomy creation into sub-tasks ( analyzing the content, describing search use cases, gathering terms as candidate concepts, etc.) and analyzes which of those tasks might be resolved with different kinds of AI technology. While generative AI may assist with some aspects, it cannot fully execute the entire taxonomy project since taxonomy features like scope notes, definitions, and non-hierarchical relationships are not easily represented in generative AI responses.

A few examples are given that illustrate this point.

According to the article, using generative AI is more suitable for various sub-tasks of taxonomy creation rather than creating a full taxonomy all at once. Using generative AI for the following sub-tasks are described:

  • Suggested narrower concepts
  • Organizing concepts into hierarchies
  • Suggesting alternative labels
  • Generating SPARQL queries

Conclusions by author:

  • Using ChatGPT/LLM technologies can help with various sub-tasks of creating taxonomies but not for taxonomy as a whole.
  • Parts of taxonomies created with ChatGPT still require human expert review to correct and refine.
  • Some taxonomy-development tasks, such as obtaining input from users through interviews and focus groups, cannot be done with generative AI.
  • Using LLMs on your own content and data will provide much better context and consistency of terminology generative AI building of taxonomies.

This article would be especially interesting for:

  • Category managers
  • B2B eCommerce Managers
  • AI Consultants

Full article text:

Subscribe to B2B Experts Digest

Contribute to B2B Experts Community

Subscribe to B2B Experts on LinkedIn