Home » taxonomy
Category Archives: taxonomy
Themes and Trends from Taxonomy Boot Camp London
“And history was the reason why she would never go to London. She saw it as dominated by the Bloody Tower, Fleet Street full of demon barbers, as well as dangerous escalators everywhere.” – Anthony Burgess, Inside Mr. Enderby
After a six year hiatus initiated by the Great Plague of 2020 and continued with a lost conference venue, Taxonomy Boot Camp London returned in 2026 at the America Square Conference Center. The event was co-located with the KMWorld Europe conference, allowing attendees to come together for keynote sessions while attending one of two tracks for each conference. Just blocks away from the Tower of London and literally encompassing part of the ancient walls of the City of Londinium, the steadfast and ancient hosted audience to the rapidly changing world of knowledge organization systems and artificial intelligence.
As with past conferences, I’m going to sum up some of the key themes and trends of Taxonomy Boot Camp as I heard them.
Working with AI
A prevalent theme across all sessions was working with AI rather than against it. While there is common concern that AI will replace jobs and remove the human from the work equation, most sessions focused on using AI as a tool to accomplish tasks that are repetitive, time-consuming, or inconsistent. In particular, using AI to identify and extract entities for taxonomy building or inclusion, summarizing large quantities of text, or automatically classifying content using taxonomy values.
Another key focus was on the probabilistic notion of machine learning models:
At a high level, probabilistic AI models uncertainty and provides outcomes based on likelihoods. This means that it doesn’t always offer one definitive answer but instead provides a range of possibilities with associated probabilities. Deterministic AI, on the other hand, is rule-based, designed to yield specific, predictable outcomes without room for variability once given a particular input. (Decision Point Advisors)
Machine learning models may generate different answers to the same question, often termed a “hallucination”. Grounding machine learning models in knowledge bases, including deterministic models like graphs in the form of taxonomies and ontologies, can create a neuro-symbolic AI approach providing more consistent answers.
Content and Training
Working with AI also means curating the training data made available to machine learning models. Publicly available large language models (LLMs) are trained on easily accessible large data sets…specifically, content available on the Internet. As we well know, content quality on the Internet is as varied as the people who create it and make it available. While LLMs then get the benefit of a variety of input, they also suffer from the biases inherent in that input. Using synthetic training data or retrieval-augmented generation (RAG) to supplement the pre-existing LLM training data can improve results. In particular, using organization-specific knowledge bases of training data can help provide more specific responses applicable to your domain with fewer erroneous answers.
Deciding what training data to use and how taxonomy and ontology structures become part of that training data is partially in the purview of taxonomists, so becoming familiar with which LLMs are being used and for which use cases will be important parts of a taxonomist’s changing role.
Language Consistency
While this is nothing new, many sessions focused on keeping the “controlled” in controlled vocabularies. Since nearly every session linked back to AI in one way or another, even the context within which we consider the basic tenant of control in a controlled vocabulary was emphasized as continually pertinent. With differences in language across one or more semantic models, machine learning outcomes are put at risk. As more areas of application for machine learning are being found, we are also venturing into areas involving more risk, like mental health or medical advice, legal enforcement, and financial decisions. Language consistency used in semantic models and as applied as metadata for training content is now more important than ever.
Among the use cases pertinent to this consistency is tagging documents at more granular levels, including inline tagging and tagging content chunks. Again, this is nothing new and has been a practice of DITA for 20 years. However, being able to consistently and accurately create training data to balance probabilistic large language models with deterministic knowledge can counter machine learning hallucinations and create more trustworthy AI agents.
“Bond”ing
And, although not a taxonomy or knowledge management theme, I did notice another commonality across at least two presentations: James Bond as an example. Perhaps it was the London venue that caused presenters to use our favorite secret agent as an example, but there he was, connected semantically to movies, sports cars, and identification numbers. I myself created a simple Bondtology for illustration purposes in past workshops and webinars. Interesting that a profession like spycraft met at the intersection of establishing deterministic truth through semantic models to avoid being deceived by artificial intelligence.
Taxonomies and the Fall of the House of Escher
“I know not how it was–but, with the first glimpse of the building, a sense of insufferable gloom pervaded my spirit.” – Edgar Allen Poe, The Fall of the House of Usher
I consider well-constructed semantic models akin to constructing a foundationally sound, well-architected, and visually appealing building. Not to be hyperbolic and melodramatic, because [swoon] that just isn’t me, has any taxonomist looked upon the works of others and despaired? Upon casting eyes upon and getting the “first glimpse of the building” that is the organizational semantic structure, suddenly felt “a sense of insufferable gloom” pervading the spirit? Boy howdy, have I.
To be fair, there are a number of factors at work leading to semantic debt; compromises in semantic integrity violating best practices in taxonomy construction left for some future taxonomist to unwind. These can be strong organizational cultural pushback to accepting taxonomies in structure or content, internal politics, or designing so that consuming systems can ingest the data. In any case, violations in taxonomy best practices can compound over time, leaving the semantic models in the current state not particularly semantic at all.
While it is mission-critical to gather input from business stakeholders to build, implement, and maintain taxonomies, it is also critical to allow the respective subject matter experts in taxonomy and business to do their work according to the best practices of their domains. Like a building drafted by M.C. Escher and constructed by Edgar Allen Poe & Associates, taxonomies can become circuitous, recurring, and not very meaningful if straying too far from best practices.
Escherian Design
I have frequently been involved in taxonomy design projects in which the stakeholder input into the semantic structures follows a line of thinking mirroring the work the business users do. In fairness, taxonomies should support whatever use cases assist end users in performing their jobs. However, business stakeholders are not necessarily taxonomists and so their recommendations may not follow taxonomy design principles. Here are some taxonomy design suggestions I have seen.
Taxonomies as virtual end caps. In this scenario, product owners try to mirror their product placement in the physical world as taxonomy structures in the virtual world. So you may get suggestions to build taxonomies like this:
Lumber > Deck building materials > Nails
Roof building materials > Shingles > Nails > Roofing nails
In essence, the concept representing the objects in the physical world are placed in the same locations in taxonomies as they would be in the layout of the store. The terms become conceptual end caps, quick items to throw in your cart because they are related to the products you are purchasing. In this case, I need nails for specific reasons, like building a deck or putting on a roof. For convenience, I put an end cap display of nails in the lumber department or by the stacks of shingles so buyers don’t need to hit every department to complete a project.
Taxonomies as navigational structures. While taxonomies can absolutely be used as navigational structures on the front end, the proposal here is that taxonomies exist this way in the back end taxonomy management system. Taxonomies may then be built like this:
Apparel > Men’s > Basketball shoes
Apparel > Women’s > Basketball shoes
From an access perspective, these are easy to understand navigational pathways leading directly to a set of products I can then filter by size, color, or brand to see what’s available but also make it easy to make a purchase.
Taxonomies as processes, stages, or funnels. Building taxonomies following process steps, stages, or trying to capture marketing user journeys through the funnel so that structures can look like this:
Awareness > Consideration > Conversion > Loyalty
Planning > Design > Prototype > Design for manufacturing > Manufacturing > Post-manufacturing
In this case, the sequential steps or stages are nested as a hierarchy as if to illustrate the progression through the process as a ladder or directional move through the concepts.
These are just a few of the examples I’ve experienced when working with stakeholders in the taxonomy design process. What’s wrong with giving the end users what they want by designing enterprise taxonomies to adhere to some of these patterns?
Maurits “Context” Escher
“If everything means everything, then nothing means anything.” Like Rick from Rick and Morty, I’m trying to build a following around a catchphrase. I made this same point in my blog Polyhierarchy and the Dissolution of Meaning. Repeating the same concept in multiple locations, whether trying to mirror real-world endcaps or to capture new contextual meanings from hierarchical placement, is a big taxonomic no-no and for good reason. If concepts become contextually dependent, then the individual subjects and objects within a semantic model lose their crisp focus. The point of taxonomies is to disambiguate concepts and ensure that each item is clearly defined in meaning and scope. Of course there are concepts that really can exist in more than one location in a polyhierarchical structure, but these occurrences should be minimal and not be forcing different contextual meanings.
Using the above examples, “nails” violates the “is a…” principle in that they are not semantic children of their parents. They are necessary items to complete a deck or a roof, but they are not decks or roofs themselves. We can easily build separate, mutually-exclusive taxonomy schemes or branches and connect them with semantic relationships to include all of the items necessary to build a deck or put a roof in place. Nesting them in contextual proximity is not following taxonomy best practices and, ultimately, causes ingestion confusion when stripped from context. More practically, repeated concepts will likely break a consuming application when the system finds the same label (and, if built properly in a taxonomy management system, same URI) showing up in two different locations. These are often ignored on ingestion because the system can not resolve the entities.
Building Codes and Accessibility
If you’ve seen any architectural drawings by Escher, you probably know that not only would they be very difficult to build in the real world even by Edgar Allen Poe & Associates, they would never pass local and state building codes for accessibility. Look at all those stairs! Not a ramp or elevator in sight!
Providing accessibility to products (or content) using navigational taxonomies is an excellent way to assist users in getting to what they are looking for. While there are more searchers than navigators in the world, simple drilldowns to products or content in hierarchies in conjunction with filters is still useful as an additional means of locating products or information. Navigational taxonomies rely on their contextual construction to provide signposts for users to know exactly where they are in the product structure and in the potentially very large “store” they are trying to navigate. Pretty self-explanatory name for these types of taxonomies.
Navigational taxonomies can be built directly in front-end applications to serve retail and information finding use cases. If possible, the values can come from back end taxonomy management systems to ensure consistent concepts and messaging across the organization. In these cases, the front end system may consume values from across the taxonomy schemes and hierarchies and display them in a different contextual hierarchy or as filtered values in left-hand navigations. It may also be possible that the taxonomy management system allows for the construction of semantic master schemes which can be reassambled in the tool or through the API into navigational hierarchies. Using our example above, the taxonomies behind the scenes may look like this
Products > Apparel > Footwear > Basketball shoes
People > Demographics > Men’s
In this case, only the values needed to construct a navigational taxonomy are pulled from their respective schemes and reassembled. The advantage to this methodology is that one best, preferred concept label and its unique ID are used in all locations. Any tagging to product images, copy, web pages, or concepts used in navigational structures or filters can be used for a variety of analytics including clicks on navigational nodes or filters, clicks on product images, analysis of products added to carts, etcetera, without having to reconcile the same or similar values for analysis.
Temporal Ladders
Taxonomy structures typically follow a parent-child “is a” structure in which the children are instances of their parent concepts. It is also possible to construct whole-part relationships (called meronymy in linguistics) in which the children are a part of the parent concept.
While it is possible to model temporal or sequential events in taxonomies and ontologies, it typically requires advanced skills in ontology modeling, can be challenging to implement, and can be subject to change when trying to mirror processes. Processes are not only sequential, but can change frequently as well. Changing a foundational semantic structure to keep in pace with changes in marketing funnels or manufacturing processes may not be worth the effort if the steps can be created taxonomies independent from their hierarchical structure.
That all said, using relationships to define sequence rather than hierarchical structure can be one simple way to create a semantic sense of order. For example, using a relationship like has predecessor could link books, films, or process steps in order to model sequence.
It’s all a Question of Time
As context graphs are gaining momentum in at least understanding if not yet implementation, we will likely see more ways to bridge the taxonomy modeling-temporal process gap. In the meantime, adhering to foundational taxonomy best practices is a best bet to ensure that your semantic models are ready for the next evolution to capture temporal events to provide additional context to the graph.
In short, maintaining “is a” or whole-part taxonomy structures as base semantic models while developing more complex ontological designs and connected data as part of a context graph will potentially provide a good combination to avoid Escherian design practices and Gothic horror in your semantic structures.
Smoke and Mirrors
“But it’s always been a smoke and mirrors game / Anyone would do the same.” – Gotye, Smoke and Mirrors
It’s pretty common for organizations to have many software systems and platforms. Some are home grown, some are commercial; some have small scopes while others are enterprise grade; some are legacy while others are newly rolled out; some are seamlessly integrated while some (um, many) are standalone and siloed. What many of these systems have in common is that they are dedicated to one or more functions. They may be for digital asset management (DAM), content management (CMS), product information management (PIM), customer relationship management (CRM), and so on. Something that few enterprise systems do well is taxonomy management. And, to be fair, why should a standalone system dedicated to a particular purpose also be excellent at taxonomy (and ontology) management? There are dedicated systems for this function as well.
Typically, taxonomy management systems (TMS) are positioned as centralized repositories for controlled values which can be integrated with multiple systems in a hub and spoke model. A centralized taxonomy architecture ensures single concept values with a unique identifier can be used across multiple systems for many use cases. A centralized architecture makes sense, but there are many challenges arising from consuming downstream systems’ inabilities to handle the rich semantic models published from a TMS. Consuming systems may not be able to ingest properties, relationships, or even hierarchies.
What are some ways we can address these integration issues while maintaining an architecture in which a TMS is a centralized source of truth for metadata values?
Smoke and Mirrors
When the first iPhone was released, it shifted paradigms. The original iPhone wasn’t the first device to include touch screens, but its form and user experience was unique in many ways. Apple is touted for their designs, and it was the combination of form and function that garnered such immediate success. We didn’t learn the iPhone, the iPhone taught us. We learned to scroll, swipe, and tap so these basic functions became ubiquitous across many devices and manufacturers. A good user experience will do that: teach us how to navigate through an application, how the functionality works, where we should expect to find a “yes/no” or “next” button, and the overall design principles. I see this as domain teaching and reinforcement.
The same is true in the world of semantics. As semantic practitioners, we must educate users and meet them at the level of their knowledge need. Some business users will only need to know the basics of taxonomy principles when a taxonomy redesign is in progress or they are onboarded as taxonomy consumers. Other business partners will want to become more involved in the process and go beyond concepts into modeling the semantics of their domain.
Modeling conceptual domains is what semantics is built for, reflecting the ways of thinking in the organization, including business processes and relationships between concepts and things. Semantic models are mirrors of organizational thinking. Consuming systems may be mirrors of these mirrors, even when they are not designed to be semantic modeling platforms themselves.
The most basic smoke and mirrors game is aligning concept label values across systems. While the values may be mirrored, it is more smoke than anything else, concealing the incredible amount of overhead in agreeing on the label values, maintaining the governance process to ensure these labels stay aligned, and workaround maintenance required by the nuanced differences between consuming systems. In these scenarios, any necessary change in any of the systems (including the taxonomy as an aligned source of truth) can be precarious as downstream values may be driving workflow processes or are hard-coded, making them extremely difficult to change.
Consuming systems supporting hierarchy, definitions, and synonyms can mirror hierarchical structures and at least some of the properties adding semantic context to concepts. Consuming systems only supporting flat lists may have the ability to create dependent fields. Although not a true hierarchy, selecting a value from one field that constrains the values in another field mirrors hierarchical structures. Again, mostly smoke and mirrors concealing the amount of semantic design it takes to understand which parent values may drive dependent fields and how the child concepts are displayed. A more advanced step toward alignment is ensuring that all of the values share a common ID, even if this ID is manually entered for each mirrored concept so they are the same across systems. Not ideal, but having a unique identifier for each concept can bring these concepts closer together even if they are pulled directly from a source of truth taxonomy management system.
More Mirrors, Less Smoke
The best way to take advantage of semantic models is to pull taxonomy concepts and their associated properties—including fields like description, scope note, and the URI or GUID—into consuming systems so there are taxonomy terms for metadata tagging and their associated attributes for context and additional information. Tagging interfaces might be dropdown lists of 10-15 values, hierarchical browsing, or typeahead fields displaying taxonomy concepts as the user enters characters. All of these methods of applying taxonomy concepts reinforce the terminology used in the organization. Dependent fields and hierarchies reinforce hierarchical relationships between concepts and reveal how ideas are organized.
Users less frequently see full ontologies including concepts and the relationships between them. However, revealing these structures in browsable visualizations can be useful to help users understand concepts and the relationships between them. Again, these semantic structures mirror the activities of the business. If users can see interrelated concepts, and preferably in the context of content tagged with those concepts, it reinforces the mission and activities of the business. Additionally, showing graph visualizations can help shift thinking from simple hierarchical structures to more semantically rich graph structures. When business users understand the current domain thinking, they can also begin to understand how these semantic structures can be applied to business use cases to solve common organizational problems. In addition to active education by taxonomists, user interfaces and mirrored domain modeling in systems can help users understand semantic modeling and what it can do for the business. Revealing semantic models can be reinforcing, but it’s probably less common and useful than more practical mirrorings in systems that don’t have particularly strong semantic foundations to begin with.
Conceptual domain mirroring takes creativity, using the functionality of the TMS, APIs, and consuming systems to create an ontological and graphical domain representation even in systems that are foundationally relational and hierarchical. Taxonomists, working hand in hand with information architects, can find opportunities to express the complexities of semantic models in existing UIs and through creatively creating new ideas meeting the use cases of the business. Mirroring domains creates organizational alignment and reinforces domain thinking without conscious effort by the users. Just as well-designed hardware and software teaches us new domains, so can we as semantic practitioners find ways to passively educate our users about domain thinking and semantics in the organization.
The AI Bot Wars
“A robot may not injure a human being or, through inaction, allow a human being to come to harm.” – Isaac Asimov, I, Robot
Someday in the distant future, automated artificial intelligence bots will wage misinformation, disinformation, fake news, and propaganda (University of Montana) campaigns directly against each other as a form of information and psychological warfare aimed at civilian populations. These campaigns will serve to erode trust, sow confusion, and create chaos within an enemy’s society. Hot wars, waged by humans or by drones and robots, will only be necessary as mop-up operations to consolidate power and assert authority. These wars will let peoples’ own interpretations and imaginations weaponize messaging against their fellow citizens until they destroy themselves from the inside. A fictionalized account of this type of hybrid warfare mixing misinformation campaigns, cyberattacks on infrastructure, and conventional military was the plot of the recent movie Leave the World Behind.
Bot wars are not the fiction of the distant future, however. They are here today and they are improving just as rapidly as the quality of artificial intelligence. Long gone are the days of blurry photos of Nessie and shaky video of Bigfoot. Misinformation created by generative AI was a key component in the Iran-Israel conflict of 2024-2025 (EDMO) and has been central to Russia’s online propaganda campaigns (NATO).
Today’s generated images and videos are hyperrealistic and can only be determined to be fake by 1) knowing the context or content to be untrue, or 2) having access to metadata which has not been tampered with. How do we combat this onslaught of misinformation? What role do semantic professionals, including taxonomists and ontologists, have in the war for truth?
Evolution of Bot Wars
Today’s artificial intelligence wars are mostly fought by people generating content. Easy access to cheaper, faster, and better artificial intelligence tools allows any user to generate new images and text rapidly with little to no skill in video or content editing necessary. Already existing content creation and social media sharing platforms have expedited and expanded the range and audience for user-generated content, real or not. Most of these platforms can’t keep up with content review and provide no mechanism for viewing the content source, including the metadata which may reveal whether the content is real or generated using AI tools. The democratization of content generation tools has meant an explosion of content (hence the term “content creators” as, seemingly, a professional occupational title). These tools have been praised for their ability to allow users to document, in real time, true events unfolding around them. These same tools allow users to document, in real time, unreal events manufactured by them with the same ease as documenting reality. Science fiction will just be fiction, the only science involved being the technical tools used to create the fiction.
I believe the next step in the misinformation wars will be an advancement in bot-on-bot directed counter-misinformation campaigns. In fact, these wars may already be happening with the number of fabricated online personas generating content responding to other comments which, in turn, may also be the product of fake online personas. Whenever one bot posts generated content, another bot will respond, countering and confusing the messaging. There may be truth in some of the counter-messaging, posting real content in direct response to fictional content. But, really, why bother with the truth at all? One bot can simply respond with equally outrageous content rebutting or retaliating against the first. Since artificial intelligence can generate content so quickly, why not take it a step further and do what any good marketer would do, segmenting and personalizing content to audiences based on their previous social interactions, including posts, likes, and network relationships. Not only can misinformation be generated quickly, it can be tailored to segmented audiences to trigger the most resonating and visceral reactions: fear, rage, mistrust, joy. Eventually, without any direct human intervention at all, peoples’ confidence in truth erodes and the reinforcement of already held beliefs and biases are strengthened. We already talk about echo chambers; the next echo chambers will be bots talking to bots with segmented human audiences receiving the exact messaging they would like to hear. Even as I talk about “will”, these trends are emerging on social media platforms today.
Recursion
I think “recursion”, “a computer programming technique involving the use of a procedure, subroutine, function, or algorithm that calls itself one or more times until a specified condition is met at which time the rest of each repetition is processed from the last one called to the first” (Merriam-Webster) is a great way to describe the more general content feedback loop we currently, and will increasingly, find ourselves in.
The referencing of original sources into various new content forms is happening increasingly in media as authoritative, unbiased news sources and are replaced by opinionated, subjective, and polarized “news” platforms. Algorithms on popular social media platforms weight toward content which has more interactions, positive or negative, and this content drowns out everything else. The number of memes and video clips I see repeated—or, rather, regurgitated—in my social media feeds gives a false impression that only a narrow range of topics are being covered. The breadth is shaved off at the long tails and only the highest middle of the bell curve is spit out into our feeds. Of course, these feeds are shaped by the content with which we interact, creating an echo chamber of reinforced, narrowly focused subject areas. Even as the overall amount of content expands exponentially, our exposure is limited to what we already think…or, rather, believe. Because belief is replacing authoritative fact. Our friends and feeds reinforce these notions; that unpleasant or dissonant facts are a matter of belief rather than any measurable objective truth.
The recursive, or regurgitative, nature of our content sources is going to have long-term effects on the bot wars. As AI bots create more and more content, they will seek out public sources of information and, eventually, feed their own previously created content into the self-guided learning models. Endless loops of self-referencing, recursive, regurgitated, manufactured information will act as the source of truth for new information; an endless entanglement of un-cited, untraceable, unverifiable information. As the bots play out their battles, the information will become so convoluted and unprovable that the only thing left will be belief. Even without the bot wars, we are finding ourselves here today. Belief over science or fact, individual belief over public sentiment, personal fictions over established facts.
The Battle for Semantics
From the early days of my career in an academic thesaurus to the present, the overwhelming mission of establishing “Truth” when so many concepts are only contextually true has haunted me. Fundamental, existential questions of being are, of course, at the heart of semantic modeling. Ontology is “the philosophical study of being” (Wikipedia) after all. As we watch truth and untruth blend into a bizarre miasma of half-truth in real-time, I wonder if other people in the semantic field feel the way I do. I have seen the frustration from scientists as they are dismissed as fraudsters somehow tricking the public into believing humans landed on the moon, vaccines can prevent disease, and fluoride is good for your teeth. Are taxonomy and ontology practitioners feeling the same level of dispirited frustration when they face the daunting task of asserting truth in a postmodern, truthless world? Will the AI bots win?
In the spirit of never giving up in the face of seemingly insurmountable odds, I offer up the following calls to action for semantic professionals which will, at least partially, address the coming AI bot wars:
- Lobby for increased use of semantic practices and technologies (taxonomies, ontologies, graph databases) in your organization. The use cases for semantics are real and can be clearly defined. The real work comes in convincing the C-suite that a rather insignificant financial investment in graph databases and taxonomy and ontology management software can indeed provide a large ROI.
- Taxonomists and ontologists need to engage directly with subject matter experts to ensure that semantic models accurately reflect the domain(s) they cover. Ongoing data ownership, quality assurance, and SME relationships should be an integrated part of the semantic model governance process.
- Similarly, semantic experts need to seek out and be involved with AI and machine learning activities in the organization. As foundational source-of-truth data for machine learning training sets, ensuring semantic models are accurate and are appropriately used in AI projects will help these projects be more successful with less risk to the organization.
- Target the most sensitive use cases. Semantic truth is the most convincing in areas in which the organization experiences risk. Find legal use cases tied to public content or product statements. Understand what risks threaten the company and which practical use cases semantic models can address.
- Design transparency into semantic models, including read-only access to taxonomies and ontologies in a variety of visualizations, so end users can understand and utilize them better. A significant part of any taxonomist’s job is helping users understand what taxonomies and ontologies deliver. Allowing end users to explore for themselves is a part of this work.
- Fight for the same transparency in content UIs in which metadata can be viewed by end users to understand the origin of the content, including whether it was generated by AI.
- If politically inclined, lobby for AI regulation and policies at the national and international level. Establishing regulations guiding the use, and particularly the transparency, of AI for all users will help to ensure that there are consistent best practices in how we implement and interact with AI and its generated content. In 2024, the European Union passed the AI Act, and more national governments and international organizations should follow suit.
- As a new technology, end users need to understand how AI works at least at the fundamental level. There needs to be more programs aimed at providing media literacy for the general public so they can learn how to identify and distinguish truth from untruth especially when it comes to AI-generated content.
- In support of media literacy and metadata transparency, publicly available AI-generated media detection tools need to be more common and easily usable by a general audience. These tools should have the ability to flag and identify misinformation for others.
The fight for truth will be partisan, political, frustrating, and even violent. We live in a postmodern world, but the death of truth will benefit those who create the most convincing and appealing misinformation the fastest. Counteracting these misinformation campaigns may very well be the last bastion of retaining democracy.
Taxonomy & Agile
“The purpose of a scrum is to restart play with a contest for possession after a minor infringement or stoppage.” – World Rugby, Law 19
In 2010, I worked as a taxonomist (and content and records management specialist) on a project in which I was embedded in a software development team dedicated to using the Agile methodology. At the time, I wasn’t familiar with Agile. I wasn’t a software developer and had never had the opportunity to work in an Agile shop. I learned a lot about Agile on that project and saw the advantages of developing taxonomies in tandem with software development using the Agile methodology. I presented on this topic at the 2011 Taxonomy Boot Camp Conference as “Agile for Managing Taxonomy Projects”.
Years later, Agile is still a widely practiced software development methodology and, at least in my work experience, taxonomists are still embedded in technology teams and work with both the business and developers in iterative sprints. Since Agile is really a set of practices which can be adopted in part or in whole, let me focus on a few aspects and how they work well with taxonomy development.
The Scrum
The origin of the term scrum comes from rugby, with the idea being that a single, cross-functional team is working together to accomplish a goal. If you watch ruby, you might see it, as this American does, as a chaotic scramble of pushing and shoving out of which a strangely shaped ball eventually is released, hopefully carried by a player. Both the idealized goal of the team and the actuality of the scrum sound a lot like a typical workplace, so I believe the idea fits. Scrum “prescribes for teams to break work into goals to be completed within time-boxed iterations, called sprints. Each sprint is no longer than one month and commonly lasts two weeks. The scrum team assesses progress in time-boxed, stand-up meetings of up to 15 minutes, called daily scrums. At the end of the sprint, the team holds two further meetings: one sprint review to demonstrate the work for stakeholders and solicit feedback, and one internal sprint retrospective. A person in charge of a scrum team is typically called a scrum master” (Wikipedia).
There are several scrum concepts which fit neatly for both software and taxonomy development:
- Determining the scope of work,
- Understanding the amount of time in which that work must be completed, and
- Working iteratively to accomplish the work and deliver a finished product.
The working taxonomist will note that taxonomy development is not nearly so neatly defined as to follow prescribed, time-boxed sprints. Especially true is that it is difficult to deliver independent, discrete taxonomy hierarchies (or full taxonomy schemes) due to the interconnected nature of multiple taxonomies bound by a single ontology. However, while restricting a portion of taxonomy development to a single sprint may be impractical, the overall guidelines of the functionality to be delivered in each sprint may be aligned with the supporting taxonomies and iterative development can still be accomplished and delivered in step with technical functionality.
In my experience, one of the most valuable features of the Scrum methodology is the daily scrums (or daily stand-ups). In these meetings, understanding what technical development is happening, and especially what changes may be occurring to deliver the right functionality within the sprint, is critical for taxonomy delivery. Separating taxonomy development from software development is one thing, but taxonomy delivery can’t be separated from functionality. If components of the taxonomy and/or ontology are essential for delivery, then it must be clear which components and how it is to be done within the software delivery. Will only taxonomy labels be delivered? What about alternative labels, definitions, or other properties? Will relationships be part of the ontology delivery and how will this be done? Does the functionality support hierarchy and, if so, is there a restriction around how many hierarchical levels? All of these are technical requirements, and, as often the case in Agile software development, the requirements and delivered work may change in the course of the sprint. Likewise, a taxonomist may determine that a new Boolean property needs to be added to the model to support the business use case. If this wasn’t scoped as a technical requirement, it needs to be considered in the functional development.
Epics, Features, Stories
The way that work is expressed in Agile is through epics, features, and stories. Epics are large pieces of work spanning multiple sprints. In an organization based on a quarterly planning cycle, an epic could feasibly span one or more quarters. For example, a taxonomy epic may be something like “Develop a Product Ontology”. Depending on the organization industry, number of products, and how many properties and relationships are involved, an epic such as this might be completed in a few quarters.
Features are the next sized work item, and these typically are significant pieces of work made up of many tasks. Following our epic example, a taxonomy feature may be “Engage with product marketers to define products and attributes”. In this example, there are clearly a lot of steps and individual tasks which need to be accomplished to realize this goal. Again, a feature may span several sprints, but it should not take longer than a quarter.
The most granular pieces of work are user stories (or just stories). Stories define the daily work and should be able to be accomplished during a sprint. Taxonomists don’t always follow sprint timelines, but they can be a good guide for defining whether a story is the right size. Again, in our example, a story could be, “Add preferred labels for 50 top-selling products”. The work is discrete, measurable, and includes some idea of how long this activity may take.
Some of the benefits of using epics, features, and stories in taxonomy development, just as in the parallel software development, are
- Scoping the size of the work appropriately,
- Defining all of the tasks necessary to complete the work,
- Stating measurable, realizable pieces of work, and
- Defining who the work is for and what it will accomplish.
A template for a user story might be something like “As a [role who wants to accomplish something], I want to [what they want to accomplish], so that [why they want to accomplish that thing]” (Agile Alliance). Following on with our example above, a user story may have a simple title like “Add preferred labels for 50 top-selling products” but then include a definition or business goal field with the statement, “As a product marketer, I want to make sure the top selling products are represented in taxonomy, so that product search and landing pages are optimized using taxonomy values for user searches.” Using a template like this makes it clear what the work is, who it is for, and what is intended to do. Although a good taxonomy should include definitions, scope notes, and editorial notes, the larger context around the intent of the taxonomy development can be captured where work is documented and tracked against software development work and organizational planning and goals. Using epics, features, and stories in taxonomy work is also a part of taxonomy governance, documenting why decisions were made and for what purpose.
Kanban
A Kanban board is a feature typically found in software supporting the Agile software development methodology. “Kanban boards, designed for the context in which they are used, vary considerably and may show work item types (“features” and “user stories”), columns delineating workflow activities, explicit policies, and swimlanes (rows crossing several columns, used for grouping user stories by feature). The aim is to make the general workflow and the progress of individual items clear to participants and stakeholders” (Wikipedia). Just as in software development, taxonomists can define epics, features, and stories at the appropriate level of work. Since Kanban boards tend to be less restrictive because they don’t necessarily need to adhere to sprint schedules, they may be more appropriate for defining taxonomy workflow items and how they fit within larger organizational bodies of work. “Work items are visualized to give participants a view of progress and process, from start to finish—usually via a kanban board. Work is pulled as capacity permits, rather than work being pushed into the process when requested. In knowledge work and in software development, the aim is to provide a visual process management system which aids decision-making about what, when, and how much to produce” (Wikipedia).
Especially in complex organizational environments involving many teams (software engineering, the business, the taxonomy team) and systems, tracking taxonomy work in a Kanban board within Agile development software can link non-technical taxonomy and ontology work directly to the technical work supporting its sandbox and production development and delivery for real-world taxonomy-based applications.
Summary
I’ve found that taxonomy development in the context of Agile works best when developing new software functionality. However, loosening some of the tenants of Agile to accommodate taxonomy development is possible. While the nature of software development and taxonomy and ontology development can be very different, taxonomists working within a technical team using the Agile methodology can reap significant benefits.
One, taxonomy work can too easily be abstracted and esoteric, losing its direct connection to business applications, business goals, and importance within complex, metadata-driven environments. Rooting taxonomy and ontology development work firmly in structured, documented Agile processes links taxonomy and ontology work directly to the technical functionality supporting its use in a host of applications. The Agile methodology helps make taxonomy work “real” and understandable.
Two, in organizations in which the Agile methodology is inseparable from quarterly planning and tangible software functionality delivery, taxonomy and ontology work is made visible, and, most importantly, measurable. How often, especially when workforce reduction (layoffs) is imminent, has a taxonomist been asked to prove the return on investment (ROI) for taxonomy work? Using Agile and documenting the work, its intent, and how the work was measured can provide a firm basis for developing metrics and use case examples of taxonomy and ontology work in action.
Finally, taxonomy and ontology development, release, and maintenance can be “squishy” work. Aligning taxonomy work with Agile provides a framework for how to develop and document work, how to measure work in progress and its completion, and how to map that work against technical development. Grounding taxonomy work in this way helps taxonomists stay focused on the immediate deliverables (please add this concept so I can make it live on the website) while remembering the greater strategic goals (develop this ontology so we can develop practical applications with real ROI). Taxonomy alignment with the greater goals and strategy of the organization helps to make the case for taxonomy’s importance in the organization as well as exactly what applications taxonomy supports.
In sum, 15 years after my first foray into Agile, its application and alignment with taxonomy and ontology work is still alive and well.
Truth & Consequences
“We live in a world where there is more and more information, and less and less meaning.” – Jean Baudrillard, Simulacra and Simulation
The art and science of taxonomy, developed by Carl Linnaeus, is a product of the Age of Enlightenment. From its outset, taxonomy has sought to neatly classify the world into named categories typically represented in hierarchical relation to each other. There is an essential human need to establish order in a chaotic universe, and the rooting of the world into scientific categories and nomenclature acts as a filing system superimposed on reality.
These foundations of taxonomic thinking proves to be both its promise and its challenge in a world which is now arguably Postmodern and dialectically opposed to the Enlightenment views of linear progress. Modern taxonomists and ontologists are still using Enlightenment tactics outside the scientific realm to provide meaning and order to information even as meaning and order collapse in practice. We have seen the flattening of truth as social media has provided platforms for democratized information, much of the content intentionally and unintentionally blurring truth and contextualized facts, beliefs, and baseless conspiracy theories. The fight to maintain (or establish) truth and order is at the heart of taxonomy, ontology, and semantic data work which has become increasingly in demand while also more challenging to establish in the shifting sands of truth.
Context
Ideally, taxonomies define a single, preferred concept in its one best hierarchical location. Such a concept is identified by its unique IRI (Internationalized Resource Identifier) allowing the concept to change preferred and alternative labels as necessary and add or subtract attribute properties and relationships. The foundational structure allows for both continuity and permanence and flexible change. A concept gains as much, if not more, meaning from its contextual structure as it does from its label.
Technology functionality permitting, a concept can then be used in different contexts for different use cases without necessarily requiring the display of parenthetical qualifiers. For instance, in a navigational context, “mercury” might be displayed under “planets” and “metals” with the only difference being capitalization. On the back end, the seemingly single concept is unambiguously identified as separate concepts which could be represented as “Mercury (planet)” and “mercury (metal)”. Using semantic standards like SKOS and RDF provide frameworks for representation for both human and machine understanding. These frameworks are the underpinnings for the Semantic Web.
Despite decades of work trying to establish the Semantic Web as the norm, how the vast amount of information on the Internet is used in practice for purposes like building large language models (LLMs) does not necessarily have to retain these semantic practices. Hence, both humans and machines can potentially misunderstand labels in different contexts if those concepts are divorced from their structures. Removing context can remove meaning.
On a larger scale, removing data and ideas from their context has the same result. While it may be easy for reasonable, educated people to dismiss nonsense, cleverly constructed conspiracy theories can be built out of decontextualized information blending facts, believable or established fictions, and belief. Conspiracy theories are promulgated by bad actors attempting to spread misinformation. The rapid growth and exponential improvement of artificial intelligence has made this even easier because factual gaps can be filled with generated text, images, and video. While information scientists may work to apply metadata to such content, this metadata is not typically visible, or perhaps even believable, for the average user. Truth as intended becomes reused but revised, a cut up pastiche claiming to mirror the original but actually undermining it.
We could presumably trace the death of source of truth documents to the transition from printed to electronic documents, but there are too many examples of charlatan print works aimed to deceive, either maliciously or for entertainment. No, the shift away from truth isn’t a shift in medium, it is a shift in paradigm in which truth is derived from the context, or lack of context, in which it is presented.
Belief & Complexity
Postmodernism (or, rather, the loose and varied set of practices identified as falling under the Postmodern umbrella) has flattened our perspectives of hierarchical truth power structures and destroyed the notion of objectivity. Everything is subjective, everything is belief. No longer can we argue science versus religion, fact versus fiction, right versus wrong. In some ways, the allowance for multiple perspectives has democratized a globalized world; in other ways, it has made it nearly impossible to declare semantic truths in a world absent of absolutes.
It is no longer enough to provide evidentiary truth in opposition to supposition and unfounded belief. Conspiracy theories in particular are too interesting, too elaborate, too fascinating to crumble under the hard light of truth. At the heart of much of belief is, ironically, complexity. Belief often stems from the need to simplify an overcomplicated world operating on sometimes unknown (at least to the believer) principles beyond explanation. Good and evil, right and wrong, a New World Order, to paraphrase George W. Bush. Dialectic oppositions like us and them, right and wrong, Heaven and Hell, black and white, and so on, exist to simplify and understand a world full of grey, somewhere between dialectical opposites. Again, ironically, these easily adopted dialectics are also easily supported and reified by the adopting mind by a concoction of contextual complexity aimed at creating new truths.
Perhaps, then, the popular rise of misinformation is mirrored in the increasingly complex models used by business organizations to represent their domains. When I started in this field, most organizations outside of complex domains like biopharmaceuticals and the like were content with hierarchical taxonomies. Now, more and more of these corporations require complex taxonomies and ontologies, especially to support machine learning use cases. The complexity of semantic models mirror the complexity of the world, and, therefore, can easily mirror the complexity of truths and untruths.
Semantics
These paradigms seem to spell doom for those who aspire to create truth using semantic models and technologies. If we are in the midst of a Postmodern paradigm in which truth with a capital “T” does not exist, can semantics continue to exist as a practice? The phrase “paradigm shift” exists for a reason…or, perhaps more apropos, “this too shall pass”. As people become increasingly unmoored from meaning and the sands continue to shift under their feet, I believe that eventually they will seek some rope to pull themselves out of the quicksand. We, as taxonomists and ontologists, are here to weave rope.
As semantic professionals, we must go back to our sources, cite them, and be sure they are visible and referenceable by those who adopt the semantic models we create. We must continue to argue for the use of carefully curated semantic models as sources of truth for machine learning training data. We must continue to hire talented researchers adept at seeking and modeling truth based on semantic rather than causal relationships. We must aim to create the most truthful semantic models we can, domain by domain, regardless of whether they are reused or adopted by other companies or consumers.
If we give up on truths, regardless of whether or not they are capital “T” truths, we give in to the bad actors and malevolent forces who expect us to swallow whole the truths as they manufacture them.
Taxonomy Blues
“I got to keep movin’, I got to keep movin’/Blues fallin’ down like hail, blues fallin’ down like hail/Hmm-mmm, blues fallin’ down like hail, blues fallin’ down like hail.” – Robert Johnson, Hellhound on My Trail
I was laid off recently from a taxonomy position I very much enjoyed. Rather than wait for the smart to wear off, the day after, I sat with a not insignificant IBU IPA and doubled down on the mixed emotions rattling through me to expound on some common issues I see in the work world of taxonomists. Many of these challenges I’ve dealt with in my blogs in one form or another in the past, but the day after a layoff hits a little differently when there’s a bruised ego and a true feeling of loss at play.
Like emails sent in the heat of the moment, posting a blog to social media hot on the heels of a personally emotional layoff is probably not the best idea. I’ve had nearly some time to let reality set in and revisit this writing. Surprisingly, I didn’t have to alter much of the content.
It’s Just Business
Layoffs happen. If you think your company loves you, you may have never been laid off. If it has never happened to you, then I am genuinely happy that you’ve not gone through the experience. Personally, I think it’s ok to believe in your company, drink the Kool-Aid (actually, it was grape Flavor Aid), and embrace their mission, goals, and strategy. I also think it’s natural and in the interest of self-preservation to recognize that any company will unceremoniously dispose of you when necessary. Their commitment to you will never match your commitment to their goals. Short of creating your own company and working for yourself building something you strongly believe in, this is always going to be the case. I enjoyed my company, believed in their mission, but also wasn’t one bit surprised when I got laid off.
From a detached, objective position, layoffs are as much a part of doing business as hiring when times are good. Layoffs can be triggered by downturns in which an organization’s revenue drops enough to merit reducing headcount. They can be triggered by well-meaning attempts at reducing redundancy and bloat. They can also be triggered by poor strategic decisions. Whatever spurs the layoffs, they are not always conducted in a strategic and thoughtful manner. Or, perhaps, there is a thoughtful strategy, but not one that will clearly bring about success.
No matter the impetus for a layoff, in my experience they disproportionately affect contractors and, because of what I can see immediately around me, taxonomists. There’s often an overlap between the two groups. When staff is augmented with consultant, freelance, or contract taxonomists, expect those people to be higher on the list when it comes to reducing headcount. The business likely doesn’t understand the role a taxonomist plays or minimizes the skill set as something anyone can do. As a seasoned taxonomist with years of consulting engagements behind me, I can tell you not everyone can just be a taxonomist. Like any proficient role, taxonomists bring unique organizational and research skills to bear. Shifting this work to the technology organization or a business domain in the enterprise is a misappropriation of work.
You’re a Taxo-what-now?
From years in the industry and still having not yet quite perfected my elevator pitch explaining my job, I can tell you taxonomy is not well understood. Cue the taxidermy jokes, financial tax questions, and, if you had the “ontologist” title, interrogations about whether you know anything about cancer. Speaking metaphorically, I do, and that is the metastasizing misunderstanding of what taxonomy, ontology, and semantic technologies bring to the table. Hence, taxonomists are not just laid off singly, but en masse, eradicating entire capabilities from organizations which likely had a long and painful path to establishing an enterprise taxonomy capability in the first place. With one swift slice of the oncologist’s scalpel, an entire function is excised with no idea of how to replace the missing connective tissue.
I’m sure many people think their job is extremely important, if for no other reason than it keeps one motivated to show up every morning. As part of justifying your work to yourself, and, more importantly, to the chain of command above you, there needs to be definitive and clear expressions of why the work is too critical to eliminate. Expressing the necessity of taxonomy work is essential precisely because it is misunderstood. Taxonomy work is often seen as simply gathering terms and putting them in lists or hierarchies, but the deep work of information science is frequently unseen. An inexperienced, self-nominated taxonomist is going to be at a loss when confronted by a dedicated, commercial off-the-shelf taxonomy and ontology system backed by a standards-compliant RDF triple store. That leaves the amateur with two options: do “taxonomy” in a simpler, alternative tool or establish a taxonomy capability in the organization starting with hiring a trained taxonomist to build the taxonomies and lead the effort to evaluate and purchase a taxonomy and ontology management system.
Having worked in many organizations which have gone from zero to taxonomy capability, it is no small task and can take anywhere from months to years. Taxonomists can start as contractors building taxonomies in spreadsheets and then either transition to a full-time role him or herself or lead the effort to hire a full-time taxonomist and bring in a tool. Regardless, the effort it takes to convince upper management of the need for a taxonomy program and the long journey to making it an essential part of the business involves a tremendous amount of time and resources. Establishing such a program and cutting it demonstrates a lack of understanding, an irresponsible waste of company resources, and a phenomenal strategic error, especially in the rising tide of machine learning and generative AI.
Foundations
It is widely understood and communicated that clear, accurate foundational data is essential for a business to create meaningful analytics, support strategic decisions, and train machine learning models properly. Despite the monumental efforts corporations put into building and maintaining clean data, it’s not all that common to see it done well. Typically, the issue is years of legacy data, all created with good intentions but frequently in conflict across disparate systems, inaccurate due to the passage of time, or made obsolete by shifts in strategic direction. To use a worn expression, there is no magic bullet to solve this problem. Migrating all that data to a data lake is time-consuming and still results in redundant, conflicting data. Creating taxonomies, ontologies, and tying data together with semantic layers and knowledge graphs also assists with creating and building foundational data, but these methods too can result in disparities.
There is no simple plug-and-play solution, but pursuing multiple strategies and bringing them together is not impossible. Data lakes serve a purpose, just as taxonomies and ontologies do. They are not either/or solutions, but AND solutions: structured, relational data working with structured semantic data and both describing semi- and unstructured content across the organization.You can have one strategy without the other…but why? There are realistic barriers to pursuing multiple data strategies, including budgets, resources, data ownership, and governance. Barriers do not preclude building strong capabilities to address these different aspects. Completely eliminating one or more of these pillars makes it more difficult for the business to execute a clear and effective data strategy, only to return to rebuild that capability at a later date at a not insignificant effort.
I’ll Be Back
There is a bitter vengeance tale in my head that goes something like this: you laid me off and now I’m back as a consultant making money from you to fix what you broke. Cyborg, guns blazing, blasting all your crappy taxonomies straight to hell. Well, not very likely, but I always win this one in my head (aside: does anyone ever lose the self-righteous conversations in their head?). Yeah, ok, so maybe I’m not back and may never be. Maybe the story runs more like someone either “discovers” that taxonomy is useful or stumbles on the skeletal, ancient remains of a discarded taxonomy management system half-buried in the earth, sorely out of date, and filled with the eggs of xenomorphs. Some face-hugger plants the seed of understanding in the astronaut and they decide the organization needs to do taxonomy stuff. Maybe someone listens and they hire dedicated roles to do taxonomy stuff. Taxonomy stuff takes off, becomes seemingly essential, and then taxonomy stuff gets stuffed in a round of layoffs. Maybe taxonomists are all just Cylons.
Anyway, bitter, wounded emotions aside, eliminating an enterprise taxonomy capability does your organization a disservice. Taxonomy is foundational to well-structured, semantic, governed data. Taxonomy data should be feeding your website navigation and search, applied as metadata to content, data, and digital assets, and providing a semantic layer to your products to power personalized experiences and recommendation engines. I’ll say it bluntly: if you don’t get taxonomy, you’ll never get machine learning. Or, rather, your machine learning models will never be optimized. If you think your generative AI proof of concept will run without taxonomy, it will…at first. Then, when scaling is the next step, expect it to fall on its face. Large language models without the context of your domain, your organization, what makes you you—that is, what is modeled in taxonomies and ontologies—will give you the bland, contextless results that can only be delivered by models that don’t get who you are. Taxonomists get who you are, but, well, they’re gone.
I said bitter, wounded emotions aside. Scrap that. I am very passionate and emotional about quality data. Nerd or no nerd, this is true. And your organizational truth is going to suffer without the expertise a semantic expert can deliver. You might say this is shameless self-promotion in search of the next gig, and you might be right. What is also right is that whether I’m the taxonomist hero who saves your disintegrating semantics or it’s another capable taxonomist, I’ll applaud the result. Because truth in data. Because waste and redundancy. Because efficiency. Because user experience. Because…
I’m not going to skewer the company that laid me off and dissect what I see as their poor strategic decisions. And, honestly, I don’t know what their strategy is or will be going forward. But, I will say this: I think it was a mistake. Not for me, not for the team I really admired and enjoyed working with, but for the greater strategy of the organization. Revenge tales aside, the company is going to feel the lack of governed, semantic data built by seasoned, professional taxonomists. There are people who remain in the organization who will carry the torch, but they’ve been hobbled by indiscriminate layoffs subjected to, unfortunately, a misguided data strategy. Maybe there are other options the company will pursue to fill the gap. Maybe taxonomy will come back someday when the stock prices are more favorable, but, in the meantime, no decent data strategy is complete without semantics.
I’ll close with a call to action to taxonomists. You already know how difficult it can be to build and maintain a taxonomy capability in your organization. Once established, make it essential. Make it foundational to data work wherever it happens. Integrate the taxonomy system into important, enterprise-wide data systems and strategies. No job is impervious to layoffs, but cementing the capability will, hopefully, help you avoid the taxonomy blues.
Taxonomies, the Eternal, and the Ephemeral
“Have you sped through fleeting customs, popularities?” – Walt Whitman, As I Sat Alone by Blue Ontario’s Shores
Taxonomists and ontologists are, quite reasonably, obsessed with the is-ness of things. We are, after all, classifiers, and what we classify must be able to conform to one or more categories. A significant factor in categorizing things is time. What is was is not necessarily what is is now.
Time-based categories impact taxonomy concepts in a number of ways, including defining is-ness and the maintenance and governance of is-ness as an ongoing practice.
Is-ness Terms, Business Terms
The subjectivity of is-ness is contextual. When navigating a website, categories are not always strictly semantic in their aboutness because our minds can fill in the blanks. If I navigate to Men’s > Shoes > Basketball, I know not to expect to find basketballs. I also know that basketballs aren’t shoes anymore than shoes are men. My mind fills in any missing words, which might be “Men’s Shoes” or “Shoes for Men”; “Basketball Shoes” or “Shoes for Playing Basketball”. In navigation, we don’t need to be so specific because, in this context, we are less concerned with is-ness than we are about navigational findability.
Even as we describe what things are, our terms may be subjective. What does it mean for a shoe to be a “Lifestyle” shoe? What lifestyle? Whose lifestyle? Similarly, what does it mean to be “Retro”? It depends on the product, the year, and the history of the item. These examples, most importantly, are commonly understood by people based on the immediate context; that is, they are time-based. Loosely speaking, “retro” tends to span 20-30 years…within the lifetime of the consumer. I wouldn’t expect to look for a “Retro” shoe and get a Roman solea (sandal) made of leather and woven papyrus leaves. Retro, for sure, but not what we commonly agree to being retro in the consumer product space.
What people know is slippery, but, somehow, we can commonly understand the difference between concepts with longevity and those which are ephemeral and trending.
The Ephemeral and The Trending
We live in a rapid age, arguably driven by a vast online sales force of young people created by capitalist organizations. Why spend countless millions and human resources on sales and marketing when pre-teens and teens can hype and distribute your product on online social media sites and sales platforms? Some of us have seen in our lifetimes the death of a salesman (that is, a door-to-door sales person) and the rise of the young, entrepreneurial sales people receiving free products, monetary compensation, and social compensation quantified by likes and follows. It’s ingenious, really.
The fast follow to fast following is the vaporous ephemerality of what’s popular and trending. But, hasn’t this always been the case? Weren’t people obsessed with trends and topics in fashion and the public sphere which were very quickly dropped and fell out of awareness in short order? Of course, but the nature of the online race to be ahead of what’s next–to be that trendsetter who identified and pushed the next big thing–is easier in an online world and has immediate social and financial consequences.
Here’s a fairly recent example from Google Trends. “Barbie pink” was a hot topic for several months around the release of the Barbie movie. The movie’s U.S. release date was July 9, 2023 (which, incidentally, is my birthday, and, depending on the source, the birth date of Nikola Tesla). Look at how neatly the searches rise to meet the weeks following the movie’s release and how that trend falls off by Christmas. Likewise, Oppenheimer was released July 17, 2023, and the trend pattern is nearly identical. And, of course, their juxtaposition as “Barbenheimer” follows a similar popularity graph.
If you produce products in pink–any products in pink–you are going to want to jump on that trend and ride the wave until it disappears. When one considers what this means, the ramp-up and execution is significant. Identify all of your pink products and create landing pages so people can find all of those pink products. Make sure these pages can be found with the search term “Barbie pink” without using the word “Barbie”, because you are likely not licensed to do so. Ensure you have enough product to deliver on the increased popularity while also ensuring you are not stuck with a warehouse full of unsold product when the trend tapers off. Logistics of this nature requires foresight, and, most importantly, the infrastructure to deliver on trends.
As for Oppenheimer, it seems there was no fire sale on atomic weapons.
Longevity
The concept of “literary warrant” is an important one in taxonomy creation and maintenance. Literary warrant is the justification for indexing or classifying based on the content of existing literature; literature, in the modern context, extended to electronic and physical writings of all kinds. When we use sources like Google Trends, we can say that this is user warrant: we see what concepts people are actually using and consider adding them to taxonomies.
Taxonomies are never finished. They constantly grow and are governed to maintain currency. Out of date concepts or phrases are deprecated or updated with newer terms. Terms of art may evolve, new areas of study may arise, or social trends may push terminology into or out of use. Taxonomists consider these factors when deciding whether a concept should be added to taxonomies. In general, the goal is to include terms that represent the domain but that also have some stability and aren’t changing rapidly.
Practically speaking, maintaining stability is important because taxonomies which are constantly in flux aren’t very useful. When terms change frequently, there is greater chance that the same or similar content will be tagged using different concepts. Additionally, frequently changing tags on content can be difficult to manage and result in sporadic and chaotic retrievability. From the end user perspective, not knowing which terms to search for or use can result in lower use of taxonomies for metadata application.
So, if trending, ephemeral concepts are useful and taxonomies with stable terms with longevity are also useful, how do we maximize the use of both?
The Ephemeral and The Eternal
For textual problems like trending concepts, machine learning (ML) models are a practical and effective means for identifying and routing terms. Using ML models against sources like user search terms on an organization’s properties, general user search terms across the Internet, user reviews, social media channels, and the like can generate terms which may be useful. Some of these terms are so fleeting, they could feasibly be tagged immediately to content for findability. As the trend wanes, the tag remains, but no longer is critical for retrieving the content. Other terms will have a longer lifespan and may be considered for inclusion in taxonomies.
The main questions to ask are
- What is the term source?
- How and why was it proposed by the ML model?
- Did the ML model compare the term to existing concepts in the taxonomy?
- Did the ML model use only exact match when comparing new concepts to those already in the taxonomy or did it also perform near match or other similarity vectors?
- How is it reviewed for potential inclusion in taxonomies?
- How does the ML model receive and process positive and negative review feedback to improve the model?
- How does the taxonomist know whether to add the term to the taxonomy or not?
Developing a process in which ML models identify ephemeral and trending concepts quickly and can route and act upon these as metadata can speed an organization’s response to trends. Human-in-the-loop reviewers can include subject matter experts for product or content tagging. Importantly for taxonomists, including them as human-in-the loop reviewers for potential candidate taxonomy concepts can help expand taxonomies and maintain their currency.
Maintaining the semantic integrity of taxonomies while also responding quickly to trending topics can improve an organization’s overall reaction to the market while also maintaining clean, quality data. Popular and timely.
Who Owns the Taxonomy?
“Ownership is not a vice, not something to be ashamed of, but rather a commitment, and an instrument by which the general good can be served.” – Václav Havel
In my experience, when a business begins building a taxonomy program, two related questions arise: where does the taxonomy program live in an organization and who owns it?
There are at least two paths that lead to these questions. The first, and the most common from what I’ve seen, is that a taxonomy has arisen organically in the organization based on a real business need requiring a solution. An example of this might be the development of a marketing taxonomy used for planning or for tagging assets such as product copy or images. In this case, a part of the organization has covered a narrow domain of knowledge and there is a recognition that it needs to expand and grow to serve the greater needs of the business.
The second, and less common, is that an organization recognizes the importance of an enterprise taxonomy where none has existed before and makes a calculated decision to start one. From the scattered remains of glossaries and metadata schemas, consultants or a hired taxonomist builds a new enterprise taxonomy from the ground up and sets the foundation for a taxonomy program. Because an organization must requisition for a consulting budget and new positions, a decision must be made as to where this position will sit in the organization and to whom this individual, or taxonomy team, will report.
Who Should Not Own Enterprise Taxonomy
Let me start by saying who I think should not own the taxonomy. Although I myself have worked in taxonomy in a technical group, I would advise against ownership by any group called Information Technology (IT) or some similar variant. In fact, I would actually be surprised if anyone in a technical organization disagreed with this assessment. Technology exists to serve the needs of the business and it is the business who should define those needs and requirements. Even when a technology organization leads the business in best practices for tooling, the business needs to define how and in what capacity the technology supports business processes and activities. While technologists such as information architects may be adept at building metadata models and schemas, including taxonomies, it is the business who must decide what values those metadata models include.
Now that I have stated the business should own the taxonomy, I’ll now go further and say that no one business domain within the organization should be the owner. Marketing should not dictate enterprise taxonomy needs, but should own marketing taxonomy needs. The same goes for any other specific domain within an organization, as any functioning company will be made up of multiple domains all working together to achieve common goals.
Where Does Taxonomy Live?
Following on the idea that no one business domain should own the enterprise taxonomy, so too should the taxonomy not live in a technology solution supporting one part of the business. While digital asset management (DAM) systems absolutely require metadata, the use case of applying taxonomy to describe assets is too narrow to act as a centralized repository for other business needs. Similarly, content management systems (CMS) are not the best place to store data that could also be described by taxonomy metadata. Using the business glossaries in data catalogs is valuable for describing the data living in or passing through that system, but is not the right tool to house business terms which should be applied in other repositories or, again, in a separate CMS. While any of these systems can house a taxonomy, none of them is purpose-built to provide enterprise taxonomy services.
As a former taxonomy and ontology management (TMS) software product manager, there is truth in the positioning of these tools as centralized, agnostic, metadata repositories for many (but maybe not all) enterprise use cases. Centralizing taxonomies in a tool allows for building enterprise taxonomies that can serve multiple use cases and multiple systems. Because the tool stands alone, it is less subject to changing business directions and domain imperatives. On the flip side, making the case for purchasing a standalone system that “only” houses taxonomies and ontologies can be challenging. I have written about this in my former position in a blog called Running a Successful Taxonomy Campaign.
So, Who Owns the Taxonomy?
An independent, centralized, enterprise taxonomy team should ultimately own the enterprise taxonomy and the TMS it lives in. The taxonomy team owns the taxonomy and ontology models they build, but what they build is always in the service of use cases defined by the business. Having a centralized team allows them to be in a position in which they can serve any and all business domains and work with technology groups to fulfill use cases in enterprise and domain-specific technologies. I’ve seen taxonomy teams reporting up to enterprise knowledge management or learning organizations which serve the same enterprise-wide function.
Some of the business use cases are truly enterprise while others may be for specific domains which in turn serve the enterprise. For example, values from the taxonomy used in navigation and search typeahead on the company’s website is where the taxonomy ROI is realized. Tagging product images and copy in a DAM serving the front end are also enterprise. The metadata from the taxonomies is used on assets which are likely going to live in multiple downstream systems and channels in which products are presented and sold.
Other use cases may be specific to a domain requiring metadata values which may or may not be shared with other domains and systems. However, centralizing these values also supports interoperability and business continuity should the domain decide to switch technology platforms. Rather than migrate metadata from the old to new system, the metadata can still be pulled from a centralized taxonomy management system using common GUIDs used across the enterprise.
The real key here is finding the owners within the business who will be accountable for the concept values, properties, and relationships needing to be maintained in the taxonomy. Taxonomists are usually generalists who can build and maintain taxonomies across a variety of domains. They are taxonomy subject matter experts, not domain subject experts (though they may become so over time). In this way, the business SMEs who know the subject matter can be accountable for adding new concepts and identifying concepts which need to be deprecated over time. The business owners are essential to the ongoing governance and sustainability of the taxonomy and, of course, are the people who know the business needs the best.
The working relationship with technology groups is the same. There are product managers owning technology platforms serving the business. Each of these tools can be integrated via APIs to a centralized TMS to consume all or part of the taxonomy and ontology graph for the appropriate use case.
A standalone, independent, enterprise taxonomy program will allow for service to any and all business domains without bias…except for those shared business goals at the enterprise level. The ability for all business domains to have ownership and stake in the shared enterprise taxonomies also allows for cross-team collaboration and innovation with shared metadata use.
Taxonomies and Turnover, or Johnny Pneumonic
“In any given culture and at any given moment, there is always only one ‘episteme’ that defines the conditions of possibility of all knowledge, whether expressed in theory or silently invested in a practice.” – Michel Foucault, The Order of Things: An Archaeology of the Human Sciences
When the movie Johnny Mnemonic came out in 1995, I often heard people mispronounce the word “mnemonic” as “pneumonic”. I speculated at the time, rightly or wrongly, that more people knew the term “pneumonia” than “mnemonic” or maybe one was simply easier to say. Strangely enough, I’ve now associated the concepts of memory and sickness because of the similarity, and confusion, of those two concepts. Maybe a pneumomic device helps you remember something…or maybe it helps your lungs function properly. Are we concerned about another pneumonic plague or a pending mnemonic plague? Who can say?
The Mnemonic Plague
If we live in a knowledge zeitgeist, it may someday later be defined as the death of knowledge. Or, more accurately, the death of the belief in knowledge and expertise in favor of opinion and belief. The results of successful misinformation campaigns include a skepticism of expertise and knowledge, feelings that opinions are equal to or supersede knowable facts, or even the inability to know anything at all. Postmodern philosophy has concerned itself with the idea that there is no objective truth, either foreshadowing the imminent knowledge paradigm or driving it. Baudrillard has predicted, quite accurately, that simulations and simulacra will become so prevalent that all meaning will be meaningless.
Add to this another sociological trend, the current workplace generational turnover from a larger, knowledgeable, older generation to a younger generation buried in and swept up by these currents of skepticism. With reality changing so quickly, and so many people being distrustful of what reality is anyway, are we on the cusp of a mnemonic plague? A plague in which no one remembers anything? A plague in which we all doubt the ability to remember it correctly? A plague in which we are convinced by others that reality is not reality, facts are not facts, and that memory is susceptible to convincingly reality-sounding alternative histories?
Johnny Pneumonic
When we began the great shift from paper to electronic documentation, the speed at which we would be able to create, store, and access information grew exponentially. Imagine a library which could feasibly contain all knowledge from human history across all countries, languages, ethnicities, and religions. An electronic library that would make the Library of Alexandria, the Library of Congress, and the Bodleian Library look like corner news stands. One library to rule them all. That happened, sort of, with the expansion of the World Wide Web. Vast troves of information were digitized and made available online even as people continued to create informational web pages and create documents which were born digital, never seeing the pulpified remains of a tree. In parallel, all the other kinds of news, information, and entertainment were also being born digitally and, in the great democratization of the Internet, anyone anywhere could theoretically have access to the ability to view and create information assuming they had some kind of device and network or satellite infrastructure.
Imagine the possibilities! Imagine how knowledgeable we could all become! Imagine knowing anything, anywhere, at any time, at the moment you need to know it! And some of that is true. What is also true is that people create patently false information reflecting their personal beliefs and agendas. There are millions of mediums, channels, and formats available to anyone to tell us what magical healing herbs will make us live longer and with fewer wrinkles; entire video libraries dedicated to documenting the horrors of vaccines and Western medicine; travel logs with erroneous anecdotes and false facts with the only verification being the person who created the content; videos of lizard-people trafficking human children; videos of directors faking the moon landing. All of this available to anyone, anywhere who wants to be an audience.
Whether an individual or a state actor, whether the content creator believes their own content or not, whether we can decipher what is true and what is not, all of this information is there for us to consume and form our own opinions about. That is freedom. That is democracy. That is the democratization and platform of eight billion voices. We no longer have to be slaves to a single narrative; no, we can write our own narratives and gather our own followers, forming fragmented communities in a new-world Splinternet in which we can all, finally, show everyone else a picture of what we had for lunch.
The Codification of Memory
I could say things were better back in the good old days, whenever those were and for whomever they were good, but that’s also an opinion, not a fact. The genie is out of the bottle. The train has left the station. The die is cast. We have crossed the Rubicon. The horse has left the barn. We now live in a world of snowclones, memechés, and information of dubious provenance. Compounding the intersection of easily creatable and accessible information, skepticism of expertise, and generational turnover is the ever-improving use of AI tools to generate very impressive, lifelike, and believable images to support textual messages. Don’t believe we faked the moon landing? Here’s a photo! Didn’t that famous Hollywood actor appear in a porno? Here’s the video!
And now for something completely different. Or, rather, here is me now getting to the point for my target audience of information professionals: what do we do in the face of all this misinformation? Our jobs…the best we can.
I’ve been working on taxonomies for over 20 years. At various times–sometimes from one year to the next and other times one hour to the next–I have experienced that overwhelming sense of doom, frustration, or hopelessness in the face of shockingly ignorant misinformation and opinions. I have long since abandoned any hope for the Semantic Web’s promise that real, verified meaning could be captured in logical, formalized, human and machine-readable structures like taxonomies and ontologies. Despite the acknowledgement, both philosophical and practical, that there is no Truth with a capital “t” and that all facts are context-dependent, I still try to build semantic structures that codify the truth of my organization at the moment we are building it.
In the context of any moment, there are truths we can build and verify. We can create preferred concepts and connect them with human-readable, sensible relationships and build taxonomies into interconnected graphical knowledge bases applied to a variety of content vetted by subject matter experts. Verified and vetted content may no longer be true in a week or a year or a decade. The relationship between two concepts may quickly become outdated. A preferred concept may fall out of use or fashion. Despite all of this, codifying knowledge in taxonomies and ontologies is not an act of futility; it is an act of capturing truth and memory at a moment. We can document these changes over time and look back over the history to see what was true versus what is true now.
These semantic structures are a way we can document organizational knowledge and pass this knowledge on to the next employee, the next generation, the next iteration of a company. Creating good historical records in semantic models serves a knowledge management function in being one way we can enable knowledge handover from one person to the next, one team to the next, one project to the next.
The organizational book of knowledge as written in taxonomies is often edited and changed, but it is still a book to which we can refer with confidence that what we tried to build was accurate and true to the best of our abilities. Taxonomies are the new black. Ontologies are the mother of all semantic structures. In space, no one can hear you taxonomize…but your work is still valuable.
Information Panopticon 