KMWorld 2018: Knowledge Management – Look How You’ve Grown!
KMWorld 2018 in Washington, DC concluded last week. Since the event combines attendees registered for the Enterprise Search & Discovery, Taxonomy Boot Camp, Text Analytics Forum, and Office 365 Symposium sub-conferences, co-located exhibitors hear an interesting mix of challenges and views related to knowledge management. Not surprising, some attendees are tasked with finding solutions to very narrow problems such as creating a labeling taxonomy to a set of documents or how to configure Sharepoint to improve search results. As an enterprise search provider in the age of AI, it’s not easy to withhold the bigger idea that, “Heh, you can get a lot more with AI-POWERED SEARCH”.
In one case the KM exec from the American Battle Monument Commission was looking to enable research by automatically applying taxonomy labels to PDFs related to international gravesite layouts of fallen service members. Granted, with a taxonomy in place, discovery of these particular documents is enhanced. But what about other details, not indicated in the documents?
A brief demonstration of Attivio’s knowledge management platform, would show how multiple facets such as location, timeline, and related subject matter could zero in and tell a larger story. What if a battle story surfaces where a hero made the ultimate sacrifice to save platoon members but was never officially recognized? With AI-powered search, pertinent sources can be indexed to know what time frame, what battle, which platoon, who was involved, what the mission was, what other details were reported, who’s surviving, and where others are laid to rest, all via one interface.
The magic of Attivio’s AI-powered search is to automatically extract entities and tag documents with metadata obtained through NLP and machine learning, then apply patented “Query-Time Join” technology to enable endless possibilities for association between different documents. This association is not dependent on a taxonomy that requires periodic updates, but rather allows the end-user to explore the relationship between different sets of data in a flexible and user-friendly way.
Additionally, Attivio can leverage a wealth of legacy knowledge or ontologies that enrich Attivio’s ability to generate metadata, follow relationships, and otherwise provide structure to initially unstructured information. For example, hierarchy and synonym relationships are discovered which enable more precise queries and ultimately, the best answers.
The bottom line here is this: If you want to get the most from your knowledge, automatically enable access to it wherever it lives, in whatever structure it exists, and apply smart metadata to it to magnify its relationships. This is the new grown up knowledge management, AI-style.