Leveraging hybrid AI to scale knowledge creation at KMWorld Connect 2021
Hybrid AI combines the strengths of machine learning and symbolic AI to extract deeper knowledge and insight from unstructured language data. As a result, NLP-based applications can be developed faster with greater accuracy and transparency.
Christophe Aubry, NA head of sales and sector strategy, expert.ai, discussed all this and more during his KMWorld Connect 2021 presentation, “Next Gen NLP: Leveraging Hybrid AI to Scale Knowledge Creation.”
Machine learning has a problem, he explained. Language is a critical type of data for organizations, and machine learning can struggle to detect nuances in language.
The choice of AI technique is a key success factor in natural language processing. A symbolic approach involves designing hand-built linguistic rules to apply to text.
Another approach is using machine learning to employ an algorithm capable of learning directly from training data. Most ML algorithms are not based on AI, the system uses statistical methods relying on recurrent patterns and textual associations to train itself.
A hybrid approach is a winning one, he explained. This leverages the strength of both ML and knowledge-based techniques to generate results.
The keys to success are identifying relevant data sources, creating access to integrated and standardized datasets, normalizing extracted data with relevant taxonomies or ontologies, and building a knowledge repository as a basis for your applications, he said.
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.