KM World is one of the leading knowledge management events, held annually in Washington, D.C. The conference theme – “KM & Enterprise Intelligence: Human or Artificial?” included presentations from vendors, consultants and KM practitioners. There are several co-located events including Taxonomy Boot Camp (where I gave two presentations1), Text Analytics Forum, Enterprise Search & Discovery, and Enterprise AI World.
Presentations from many speakers are posted on these sites:
I attended more than 30 presentations over the four-day meeting across these events, talked to vendors in the enterprise solutions exhibition and met with colleagues during coffee breaks, lunches and dinners.
Some of the key themes that emerged at KM World 2024 were:
- While artificial intelligence (AI) is breaking down bottlenecks in creating, tagging and using information2 the challenge is how to effectively keep humans in the loop to review and validate content, metadata and the analytics.
- Data quality is the foundation for successful AI applications3. Templates and other content engineering techniques such as componentization can help to ensure data and content quality.
- Governance is key for sustainable KM applications and programs.4
- Artificial intelligence is creating realignments in the knowledge management industry, but generative AI (Gen AI) does not replace existing information retrieval (IR) technology. Gen AI is a new programming interface, generates new content as the result, and works in combination with other information technology.
David McComb, Founder and CEO of Semantic Arts gave an excellent presentation about knowledge graphs. He pointed out that there are mostly entities (people, organizations, locations, events, and things) and the relationships between them in knowledge graphs, but “mostly no knowledge in knowledge graphs.” In this context, a knowledge graph is an ontology which is to be distinguished from a taxonomy which is a list of values for a specific type of entity. Gist is an upper level ontology developed by Semantic Arts that covers typical business concepts (functions and sub-functions). The latest release of Gist is available as a free download.
Artificial intelligence has created realignments within the knowledge management industry including applications such as enterprise search, content management, document management, learning management, collaboration, business intelligence, expertise management, customer service management, etc. as well as generalized functions such as taxonomy management, automated categorization, natural language processing, knowledge bases, etc. There have been notable mergers and acquisitions among some of these players. The rationale for this was heard in the exhibits and in many of the KM World presentations. It was frequently repeated that generative AI (Gen AI) does not supersede other methods such as the vector space model for information retrieval, named entity extraction, pattern recognition, language understanding, translation, etc. The major shift is the ability to generate summaries that are really well written, instead of just a list of results, and to use natural language as the user interface instead of the laborious process of writing computer code. The problem is how to marry Gen AI with the other information retrieval and information processing methods and techniques. I heard presentations and talked with the following vendors who all said they are experimenting with Gen AI to learn how it can be used in conjunction with their existing products be they search engines like Coveo, eGain, or Squirro (now merged with Synaptica a taxonomy management system); search services like Northern Light; or graph databases like OntoText (now merged with Semantic Web Company known for the PoolParty taxonomy management system into a new company called GraphWise).
Footnotes
- “Using Taxonomy to Move from Paper to Knowledge Graphs” and “Updating ISO 25964, Standards Process, Related Standards, and Use Cases”.
- “Once you write something down it is no longer knowledge, it’s information.” Rebecka Isaksson, KM Expert and Founder, KnowFlow Value (formerly at Microsoft)
- “Trust + ease of use = user adoption.” Ashu Roy, CEO, eGain.
- For example, see: Richard Huffine, Assistant Director, Enterprise Information and Records, FDIC (Federal Deposit Insurance Corporation) Mastering Metadata With a Data Catalog and Linda Lavelle, Founder, Knowledge Dynamics (formerly at Johns & Johnson Supply Chain) Why Good Governance Matters – Ensuring the Best Knowledge for Your Organization.