While it is fair to say that search is still a work in progress, there has, in fact, been considerable progress. Increased computing power and more sophisticated analytical techniques have elevated search performance just in the past couple of years. In particular, semantic search has allowed for much more flexibility in how a question or request is phrased. Thanks to natural language processing (NLP), large language models (LLMs), and retrieval-augmented generation (RAG), the door has been opened for systems that can interpret the meaning of text, thereby discerning the user’s need or intent and providing a more relevant answer.
Knowledge-Intensive Businesses
Healthcare and professional services are knowledge-intensive industries with a high need for storage and retrieval of critical information. Semedy’s Knowledge Management System (KMS) focuses on these two industry sectors. The system stores knowledge assets and their semantic relationships; each knowledge asset is stored as an entity. The unified KMS incorporates many KM functions, including taxonomy management and enterprise search, into one platform.
“One of our key strengths is the system’s ability to maintain referential integrity of the knowledgebase as the content evolves,” said Charles Lagor, principal biomedical informaticist for Semedy. “Corruption of the knowledgebase is prevented through a variety of entity services,” he continued, “and can be extended with domain-specific custom constraints through validation rules."
As an example, electronic health record systems contain thousands of rules that are intended to ensure the safe prescription of drugs. “These rules depend on official code sets that are released monthly,” Lagor explained. “Any change in a medication code can have a downstream impact on the medication rules.” Semedy’s system manages those dependencies so that they align with the changing codes.