Cognitive Architecture for Process Industries - Using Large Language Models for Data Analytics

Watch a 23-minute conference talk from the eSAAM 2023 event where Peter Bednar, Associate Professor at the Technical University of Košice, presents a Cross-Sectorial Big Data Processing platform featuring semantic modeling tools and automated data analysis script generation. Learn about the platform's cognitive component that leverages Large Language Models (LLMs) for automatically extracting task definitions from narrative problem descriptions. Discover the evaluation results across five different domain problems demonstrating the promising potential for fully automated data analytics in process industries. Sponsored by HE-CODECO, EU-CLOUD-EDGE-IOT, META-OS, NEPHELE Project, and the Eclipse Foundation.