As much as we like to celebrate success and great work at Taxonomy Boot Camp London, we also find value in talking about the times when things don't go so well on our projects. Practitioner of the Year award winner at Boot Camp 2019, Ed Vald highlights many pitfalls from his own experiences, including but not limited to, requirements creep, lack of user adoption, and lack of management buy-in. He offers hope for everyone who might be going through a rocky project.

A case study that begins with a 24 month journey to design and build an enterprise ontology, incorporating multiple internal and external cross-walked taxonomies, and ends with a healthcare content search and discovery success story. The Royal Pharmaceutical Society (RPS) is a world leader in the safe use of medicines. In 2020 RPS designed, developed and deployed an enterprise ontology to standardise the classification and terminology used to identify and describe chemical substances, including drugs, excipients, poisons, herbals, etc., and their therapeutic uses, chemistry and physical properties.

This session, co-presented by RPS and Synaptica, discusses the knowledge modelling design process, illustrates some of the practical workflows used by the taxonomy editorial team, and describes how the resulting ontology, SPINE (Substances their Properties, Identifiers and NomenclaturE) integrates with content and search systems.

Selecting and collecting content is a fundamental information activity, and any taxonomy project should fully consider the needs of its users.  The rise of natural-language processing techniques and statistical tools make it increasingly possible for users to make a taxonomic choice without needing to be a taxonomist. If label choices are well designed, it may be sufficient for the user to identify a solution they like and opt to see “more like this”.

Michael sets out some examples, good and bad, of this technique, where users are expected to make decisions based on an underlying classification system, and takes a look at how Google provides a means of simple selection in an imperfect information universe.  In this useful beginner's guide, attendees will learn to start questioning whether the choices they present to their users are necessary, and how to simplify the road to solving the problem.