Creating compelling search experiences at KMWorld Connect 2021

Artificial intelligence, machine learning, and knowledge graphs are changing how search is implemented and delivered.

At KMWorld Connect 2021, Joseph Hilger COO, Enterprise Knowledge LLC and Neil Quinn, senior consultant, technology solutions, software engineer II, Enterprise Knowledge LLC explained how these technologies impact search and shows examples from work done with clients around the world.

“Search is a critical piece of making knowledge management work,” Hilger said.

Artificial intelligence is an integrated system that can perform actions that traditionally require human intelligence, Quinn explained. Machine learning is one type of AI system, in which a machine “learns” from your data.

Machine learning can be used to dive into clustering, inferred relationships, auto-classification, predictive analytics, recommendation engine, natural language processing, and image and shape recognition.

Machine learning can pair well with knowledge graphs, Quinn said. A knowledge graph is can infer the relationship between things.

Good conditions for ML include access to human SMEs, well defined use cases, rich data, and tolerance for inaccuracy.

“The machine has to learn and in order for it to learn, it needs someone to teach it,” Quinn said. “If you don’t have subject matter experts, you’re going to have a hard time.”

Challenging conditions include the expectation of full automation, the desire to use “new tech,” sparse data, and the high need for accuracy.

AI initiatives fail because there’s a lack of clear business applications, the assumption that AI is a “single technology” solution, there’s a lack of human resources, and enterprise information and data is not ready for AI.

Applying AI to search can be done through natural language search, categorization and classification tools, advanced analytics, or content aggregation, Hilger said.

“The most valuable search results are aggregated pieces of information about a customer, an employee, a product, a service, and a topic within the organization,” Hilger said. “I want a result that shows me an aggregated view of everything about that.”

Use natural language to find content quickly and efficiently across systems, he explained. Chatbots can use knowledge graphs and machine learning to work with users to find helpful information and refine searches.

When an ontology’s entities and relationships are stored in a graph database and integrated with a search engine, it enables an organization to search and facet based on the relationships of those entities. When querying for content, facets dynamically populate based not only on the tags, but also defined and inferred relationships of entities. This is the foundation of Natural Language Searching.

Automatically categorize content, using a knowledge graph for context. Knowledge graph enhanced search encourages discovery of new results and related content.

Enterprise 360 uses a knowledge graph to aggregate information about people, places, and things into a single graph that understands what these things are and how the information assets fit together.

“Now what people are starting to say is they want a full picture of everything, give me enterprise 360,” Hilger said.

Machine learning mixed with search can create topic pages, which can collect information from various locations and display it in one location for users to access. It can support information discovery and browsing patterns through the aggregating of topically similar information in a single space.

KMWorld Connect 2021 is going on this week, November 15-18, with workshops on Friday, November 19. On-demand replays of sessions will be available for a limited time to registered attendees and many presenters are also making their slide decks available through the conference portal. For more information, go to www.kmworld.com/conference/2021.