Knowledge graphs and ontologies have become AI’s latest centerpiece, but for many of us, this work started way before the hype. Drawing on 20 years of experience spanning research, startups, products, and enterprise implementations, Sequeda shares the scars, surprises, and lessons learned from pursuing the same problem across changing technologies and organizational realities. Hear about recurring patterns, such as why efforts succeed and fail, why organizations consistently underestimate knowledge work and get stuck, and how teams can establish semantic foundations without boiling the ocean. If you're building, buying, or betting on knowledge graphs today, this talk is a practical guide grounded in experience, failure, and persistence.

This introductory taxonomy tutorial covers key concepts to get you up-to-speed for the rest of the conference or helps prepare you to take on a role in a taxonomy project. Topics include what taxonomies are, the various uses and benefits of taxonomies, comparisons of taxonomies with other knowledge organization systems, standards for taxonomies, taxonomy creation research and sources for concepts, the wording of concept lapels, best practices for creating hierarchical relationships, the structural design of hierarchies and facets, and brief introductions to other issues related to taxonomy management.

Domain models map relationships between information objects and reveal critical taxonomy ecosystems. They identify taxonomy needs by cataloging where and how taxonomies support content, systems, and workflows. Using a case study on re-architecting the marketing website for a major technology company, Stringer demonstrates how early domain modeling creates long-term alignment across taxonomy, information architecture, and content modeling. Together, these efforts establish a foundation for scalable content architecture, support cross-workstream coordination, and strengthen semantic consistency across digital experiences.

AI increases demand for taxonomies, ontologies, and other semantic models and changes how organizations design, maintain, and use them. Drawing on real-world examples, Griffin examines three shifts shaping modern taxonomy practice. First, semantic models provide the meaning, context, and structure that support successful AI initiatives. Second, taxonomists now design for both human users and machine consumers rather than for findability alone. Third, AI can accelerate taxonomy and ontology development when practitioners apply it thoughtfully and provide the right inputs. Griffin explores how taxonomists identify the semantic models best suited to different use cases, balance human and machine needs, preserve expert judgment, and demonstrate the value of semantics in an increasingly AI-driven environment.

AI can accelerate taxonomy creation and tagging, but without governed definitions, organizations risk inconsistency and loss of trust. Hear how OpenSesame established authoritative definitions for more than 4,000 skills in a customer-facing catalog after identifying gaps in how concepts were understood and applied. Kaari describes a human-driven, AI-assisted approach that grounds definitions in real content usage, systematically develops and differentiates concepts, and maintains quality through expert review. Learn what governance structures, ownership models, and documentation practices support long-term consistency and prevent definition drift.

Managing a product catalog is a constant battle against data chaos. As inventory scales, maintaining a clean, intuitive product taxonomy through manual curation becomes impossible, leading to misclassified items, broken search filters, and lost revenue. Learn how machine learning and LLMs can help analyze unstructured data to automate product categorization and attribute extraction with high precision. AI uncovers hidden consumer search trends to optimize category hierarchies in real time, ensures seamless cross-border and omnichannel marketplace mapping, and automates continuous data governance. Through real-world case studies, Schweizer demonstrates how shifting to an AI-augmented taxonomy framework slashes operational overhead, accelerates time-to-market, and improves the digital customer experience.

Enterprise taxonomies span organizational boundaries, yet every taxonomy function requires clear ownership and governance. This master class examines centralized, federated, and hybrid ownership models, along with governance committees and other structures that balance authority with flexibility. Drawing on implementations across large public and nonprofit organizations, Jenkins explores how organizational placement within areas such as IT, KM, or communications influences priorities, funding, and long-term sustainability. Learn about the strengths and limitations of different approaches, frameworks for assigning responsibility and building cross-functional governance, and common challenges such as organizational restructuring, staff turnover, and shifting budget priorities.

Taxonomy design and information architecture require input from multiple perspectives, but engaging stakeholders and guiding groups toward productive outcomes can be challenging. Degler and Brown examine two complementary skills that strengthen collaborative taxonomy work: facilitation and workshop design. Degler focuses on managing group dynamics, maintaining momentum toward consensus, diffusing tension, and creating an environment that supports constructive participation while balancing detailed outcomes with the tone of the room. Brown explores practical approaches for designing interactive activities that deepen engagement with taxonomy and information architecture concepts. Learn frameworks for planning and facilitating multidisciplinary workshops that encourage collaboration, improve participation, and support effective decision making across teams and stakeholders.

Modern SKOS taxonomy tools provide built-in reports for orphan detection, cycle detection, and structural validation, but taxonomists often need answers beyond vendor-defined templates. Colvin and Hedden offer a practical guide for using SPARQL, demonstrating how it functions as a custom reporting layer for SKOS taxonomies, enabling practitioners to identify missing definitions, scope notes, alternative labels, polyhierarchies, and other structural or content patterns. Using query examples, they show how SPARQL supports taxonomy stewardship, quality assurance, and governance by allowing organizations to ask questions specific to their own models and user needs. The approach assumes familiarity with SKOS and curiosity, but no prior SPARQL experience is needed.

Taxonomies, reference data, and master data each serve distinct purposes in theory, but organizational realities often blur their boundaries. Taxonomies and ontologies now support not only navigation and tagging, but also workflows, process automation, and semantic interoperability. Reference data may exist in relational databases, function as taxonomy-managed vocabularies, or integrate with master data management platforms. Likewise, taxonomy values may operate as enterprise data products distributed across systems. Lehnert examines how organizations characterize these overlapping data types, determine where and how they should be managed, and establish governance across teams with different responsibilities and priorities. Hear practical strategies for aligning taxonomy, reference data, and master data practices within enterprise data governance frameworks to support consistent, effective data management at scale.

AI entity extraction tools scale content tagging, but extracted entities rarely align with controlled vocabularies, preferred labels, or taxonomy structures. Using a case study from a technology organization managing a large corpus of marketing and technical content, Geist examines how taxonomists bridge the gap between raw AI output and taxonomy requirements through synonym mapping, preferred term alignment, and term disambiguation. The work highlights the challenges generic language models face with jargon and product names and demonstrates a human-in-the-loop review workflow that builds trust in AI-assisted tagging while improving model performance over time. Hear candid lessons about what failed, what surprised, and what worked.

Taxonomies and ontologies serve complementary roles in semantic modeling, yet practitioners often treat them as separate disciplines or view ontologies as replacements for taxonomies. This session explores how taxonomies and ontologies work together across many real-world use cases. Drawing on years of practical experience, Petrillo presents a methodology for incorporating taxonomy effectively in ontology development. Learn about guiding principles, modeling patterns, visualization and code snippets.

The driving purpose of taxonomy has moved beyond content organization to a critical lever for strengthening AI reliability and cost efficiency. For taxonomists, understanding the transition from metadata schemes to enterprise ontologies is no longer optional; it is a vital career accelerator. Explore how to overcome challenges, engage in strategic conversations, and build the business case for semantic modeling. Stroker examines how AI frameworks such as GraphRAG and MCP use taxonomies as execution maps for autonomous agents and how to "weaponize" your current metadata skills for improved AI alignment. Cantrell and Farner present case studies and practical examples to illustrate how to unlock knowledge graph modeling and GraphRAG capabilities, even while constrained by less mature information architecture infrastructure.

Taxonomies and ontologies codify concepts and relationships, yet language continuously shifts through context, culture, and organizational dynamics. As AI systems increasingly rely on natural language prompts and enterprise semantic models, taxonomists face new challenges around ambiguity, interpretation, and meaning. Lehnert examines the tension between semantically rigorous structures and fluid language use, including how context, intent, and organizational politics influence terminology and categorization. Busch draws parallels between early web search adoption and the rise of LLMs, focusing on how taxonomies support more effective AI prompting and machine interpretation. Learn approaches for designing semantic models that balance rigor with linguistic flexibility to support trustworthy AI and enterprise knowledge systems.

Join us for the opening of the Enterprise Solutions Showcase to explore the marketplace and connect with the community. Discover cutting-edge products and services from the industry’s top companies while enjoying live music, a complimentary beer and wine bar, and a selection of light hors d'oeuvres. Whether you're looking to scope out new tech or reconnect with industry peers, this vibrant reception is the perfect place to build your network.

Join us for a night of friendly competition! Whether you're a seasoned strategist or just looking to unwind, we'll have a variety of games ready to play. Come grab a seat, meet fellow attendees, and let the games begin.

The workforce is greying. Across the OECD, the median age has climbed past 44, retirement waves are accelerating, and decades of hard-earned tacit knowledge are quietly walking out the door of every enterprise—often with no plan to capture it. For KM leaders, this is not just another demographic statistic. It is the moment this function moves from supporting role to center stage. Drawing on his new book, Pontefract shares the central framework of his research: the Wisdom Wheel. Built on five components—Mentor, Collaborate, Capture, Renew, and Purpose—it is a practical playbook for what enterprise learning, human-centric AI, and knowledge sharing must look like when half the workforce is over 50. He discusses a new concept, Rivers, Rocks, and Rubies, his framework for the three wisdom-sharing eras across every team member’s career, and the real organizational challenges piling up as the Rubies retire and expertise bottlenecks, leadership vacuums, and the collapse of institutional memory runs rampant. Expect golden nuggets and cases from BMW, Deutsche Bahn, TELUS, Tata Chemicals, and Turing AI and a candid look at how AI-enabled cognitive task analysis is finally helping organizations capture the expertise that has always been invisible. How can grey turn to gold? Come find out from our always entertaining and energizing speaker!

Most enterprise AI deployments share a quiet structural flaw: They are stateless. Each interaction begins from scratch, expertise evaporates between sessions, and the 80% of institutional knowledge that lives as tacit judgment in expert intuition, cultural norms, and unwritten heuristics—remains entirely invisible to the system. Clarke provides a different design paradigm, one that treats enterprise AI not as a collection of applications but as a cognitive operating system: combining the language fluency of LLMs with the formal reasoning guarantees of knowledge graphs and a compilation layer—drawing on Karpathy's LLM wiki concept and Polanyi's theory of tacit knowledge—that converts human experience into structured, compounding, queryable intelligence. The organizations that lead the next decade will not be those that find the best model. They will be those that build the best memory. Grab lots of insights and ideas from our experienced and popular speaker.

Enterprises are racing to deploy AI agents, but most are feeding them the same messy content that has confused human agents for decades. The result is predictable: Confident answers that are wrong, inconsistent, or out of policy. Gopal makes the case for a fundamental shift in how we think about KM: not as a library of content for organizations to store and humans to search, but as a set of instructions for AI to follow. When knowledge is structured, governed, and trusted, AI progresses from carefully parametered pilots to confident and consistent actions at scale. When it isn't, no amount of model horsepower can save you. Drawing on real-world deployments across banking and financial services, telecom, government, and healthcare, Gopal shows what it takes to turn institutional knowledge into instant, trusted answers at scale and why the organizations that treat knowledge as instruction, not documentation, will be the ones whose AI actually works.

This fast-paced keynote analyzes current data trends, including the limits of vector search and the traps of un-governed, auto-extracted graphs—to explore what true semantic grounding requires. Stripping away the hype, Stroker takes an honest look at the time, budget, and cross-functional effort needed to build an interconnected knowledge graph, counterbalanced by the tangible payoff: a deterministic, future-proofed cognitive asset that drives real organizational knowledge. Get lots of ideas and insights from our enterprise data strategist.

Building a taxonomy for audiobooks sounds straightforward until you realize the domain spans centuries of publishing history, multiple classification traditions, and a user base with wildly different expectations about what a "category" means. Buckenwolf traces the evolution of Spotify’s audiobook taxonomy from a manually developed model through a major redesign and into an AI-assisted workflow managed by a small team. Hear about what breaks when handcrafted models scale, how AI accelerates taxonomy redesign and evaluation, and where AI-assisted approaches fall short. Drawing on practical experience, Ruiz shares lessons on taxonomy evolution, governance, and sustainability and explores how taxonomists can design semantic models that support discovery today while serving as more effective inputs for future AI applications.

Successful AI outcomes depend on the quality and structure of the knowledge assets that support those systems. Lee and Glerum examine how Adobe prepared for future AI capabilities through a pilot focused on improving knowledge assets using KM practices and semantic design. Their work incorporates taxonomies, ontologies, and business glossaries to strengthen the organization’s technical landscape and support AI-driven employee experiences. They demonstrate how AI-readiness efforts improve the quality and accessibility of enterprise knowledge, reduce hallucinations, support more accurate responses to complex questions, and increase AI adoption. The case highlights how KM and semantic technologies provide the foundation for effective AI implementations and help organizations secure stakeholder support for long-term initiatives.

With over 85 years of storytelling history, Marvel Comics has created a universe with over 30,000 issues, thousands of artists, a huge cast of characters, and a wide variety of data that needs to work together just as well as the Avengers. As technology and data continue to evolve, there can be challenges in the comic book industry to design an ecosystem connecting all aspects of a growing universe. Hochstein and Petrillo explore what they considered while building a knowledge graph data model to create deeper semantic connections across stories, projects and asset management. Hear how they assessed a wide range of use cases to describe story content, adding richness, context, and integrating it with other important data building blocks.

GenAI is rapidly changing how organizations create, maintain, and apply taxonomies. In this panel discussion, practitioners from financial services, professional networking, media, educational publishing, and AI technology share their experiences using LLMs and other AI techniques to support taxonomy and ontology work. Panelists discuss applications such as term mining, taxonomy maintenance, genre and category development, structured knowledge creation, metadata enhancement, domain coverage analysis, and machine-readable knowledge representation. Hear about practical successes, implementation challenges, lessons learned, and emerging approaches for integrating AI into taxonomy workflows. Together, the panel explores how AI is reshaping taxonomy practice.

2:14 a.m.: An enterprise AI assistant answers a routine executive question. As the workday starts, the response triggers a compliance review. The answer appears polished, complete, and grounded in enterprise knowledge—but it is wrong, and no one immediately knows why. The “Knowledge Heist” turns this failure into an investigation. Participants reconstruct how a GraphRAG-powered system arrives at a confident but incorrect answer by piecing together evidence packets that include taxonomy excerpts, synonym behavior, relationship mappings, and retrieval traces. Along the way, uncover how seemingly small decisions in taxonomy design, content tagging, graph relationships, and governance combine to shape AI outcomes, and how stronger governance and graph integrity support more reliable enterprise AI.

The taxonomists have done their homework, now it’s time to face the Sharks. Throughout the conference, we will crowdsource ideas inspired by some of the biggest challenges, opportunities, and “what if?” scenarios in the world of taxonomies, metadata, and AI. Join forces with fellow attendees to refine the most compelling concepts before presenting them to a panel of Semantic Sharks for feedback and investment-worthy consideration. Which ideas are visionary? Which are practical? Which are just crazy enough to work? Expect spirited discussion, creative thinking, and a celebration of the ingenuity and imagination that drive our profession forward.

Wind down after a full day of stimulating sessions with a casual happy hour right on the showcase floor. Grab a drink, visit the booths, and dive deeper into conversations with fellow attendees, speakers, and sponsors. It’s the perfect, laid-back setting to talk shop, swap ideas, and solidify your new connections.