Using knowledge graphs and search - KMW Webcasts KMWorld Live
With newer technologies, the digital workplace can dramatically improve employee engagement, benefit from data-driven decisions, and enable actions that serve tangible business objectives.
Recently, Justin Sears, vice president of product marketing, Lucidworks, and Karl Hampson, director of artificial intelligence, office of the CTO, Solstice, provided a deep dive into knowledge graphs and where they fit in the machine learning landscape during a KMWorld webcast.
Knowledge graphs create a human-readable network of facts, Sears said. Knowledge graphs describe real-world entities and their relationships such as objects, places, and events.
People interact with knowledge graphs every day, when they use Google, Facebook, and LinkedIn, though they may not be aware of it. In 2012, knowledge graphs became more prominent when Google introduced the “Things not Strings” knowledge graph application, and now there are many solutions and tools emerging.
Knowledge graphs and search go hand in hand, according to Sears and Hampson. A search index contains many mentions of things, people, places, and events with implied relationships. A semantic knowledge graph extracts and navigates them dynamically.
In a traditional setting, natural language processing can help analyze text to populate knowledge graphs automatically with facts in the form of subject-predicate-object, explained Hampson. This can be a highly complex process, and the barrier to entry is high.
A semantic knowledge graph of conceptually similar things is derived automatically from indexed data by analyzing the implicit links between the “things” across the corpus. The benefits of semantic knowledge graphs are that they move beyond matching just words in searches to things and related things, automatically derive insights by discovering connections and providing “thing-based” navigation of unstructured data, provide automatic recommendations, and enable predictive analytics to help users understand the importance of things and the likelihood of outcomes.
Semantic knowledge graphs can help connect users and employees to the most relevant information they need to make better and faster decisions in the real world, Sears and Hampson said.
A list of archived webcasts and upcoming web events is available at www.kmworld.com/Webinars.