Abstract

Taxonomies have played an increasingly important role in information systems research since the publication of Nickerson et al. in 2013. That paper, sometimes called the “seminal work” on the subject (Kundisch et al. 2022) with 2000 citations according to Google Scholar, presents a method that has been called the “de-facto standard” for developing taxonomies (Kundisch et al. 2022). We estimate that over 700 papers have developed taxonomies using this method. However, that paper is not without its limitations. For one, the method described in the paper (called the NVM method after its authors) applies almost exclusively to faceted (that is, “flat”) taxonomies but taxonomies can also be hierarchical (e.g., Weking et al. (2020)). Nickerson et al. 2024 partially addressed this issue. Still other issues in taxonomies and taxonomy development have arisen. This TREO talk will outline some of the issues that need to be addressed for a full understanding of taxonomies and taxonomy development in information systems. The following list presents our first attempt at outlining some of the important issues. We encourage others to add to this list: 1. What is a taxonomy? Is it a model, method, theory, artifact, something else, or any of these depending on the circumstances, or perhaps a combination of these? 2. If a taxonomy is a theory how can it contribute to the body of knowledge in information systems? If it is not a theory, could it lead to a theory in information systems? 3. What values do taxonomies add to IS research? Under what circumstances should a researcher consider developing a taxonomy as part of their research? 4. Why is there a proliferation of taxonomy papers? 5. Is the NVM method still valid or does it need to be modified or updated, or even discarded and replaced with an entirely new method? 6. How do we develop a hierarchical taxonomy? Does the NVM method apply to hierarchical taxonomy development, or can it be adapted for hierarchical taxonomies? 7. Can two or more taxonomies be merged or integrated to create a new taxonomy? 8. Can a taxonomy be split into two or more taxonomies to create new taxonomies? 9. How can we validate a taxonomy? How do we know if a taxonomy is “correct” or “good”? 10. Is AI an alternative/better approach to taxonomy development? What problems arise in applying AI to taxonomy development? In this TREO talk we will briefly discuss these issues and encourage the audience to raise other issues in taxonomies and taxonomy development. (References are available upon request.)

Download

Share

COinS