Abstract

Ideally, knowledge about information systems should be identifiable and accessible by practitioners who might use that knowledge and researchers who might study and improve it. Related challenges are implied by a paper (Hassan & Mathiassen, 2018) that applied a “distilling” process to 6643 articles in the Basket of Eight between 1978 and 2012, thereby identifying “466 ISD [IS development] articles that offer canonical ISD knowledge distinctive to IS and complementary to other disciplines.” While the 466 articles are categorized into meaningful categories, a scientific or engineering discipline should aspire to providing knowledge in its domain in a form that is more accessible and usable than the content of 466 articles, many of which are surely inconsistent or obsolete. More important, the articles are containers of knowledge and in combination do not support easy access and use of that knowledge even with the help of current LLMs. This one-page TREO paper identifies steps toward organizing IS knowledge for accessibility. The goal is to organize knowledge about information systems, not knowledge about IS as a discipline, about doing IS research, about philosophy, etc. The first three of the six steps below are based on a stream of research that builds on Alter (2018), which was clarified and extended by Alter (2024, 2025a, 2025b). First, identify a basic unit of knowledge. This can be deemed a knowledge object (KO). The purpose of science is to identify, evaluate, disseminate, and improve KOs. Second, propose a taxonomy of KOs. An evolving taxonomy of KOs updated in Alter (2024, 2025a, 2025b) includes five categories of KOs and types of KOs within each category. The categories are concepts, data, interpretations, generalizations (e.g. models, principles, theories) and methods and tools related to creating, operating, maintaining, and improving ISs. Third, clarify how to describe and evaluate KOs. Alter (2024) proposed eight evaluation criteria. Fourth, identify and organize a large set of KOs that might be useful for practitioners and researchers. Results of initial explorations imply that a first draft can be produced with the help of LLMs. E.g. using prompts that ask for 25 prominent IS theories, 25 IS security principles, 25 IS phenomena and so on. Ideally, developmental iterations of this step would involve cooperation between researchers who would identify overlaps, omissions, and ambiguities that need to be confirmed or resolved on the path toward a large set of useful and internally consistent KOs. Fifth, compile the KOs from the fourth step in a knowledge graph (like the vastly larger Google knowledge graph used in Google search) that can be accessed and used through an interactive tool. Sixth, test the value of the result through use in teaching, research, and practice.

Comments

tpp1321

Download

Share

COinS