This talk presents an introduction to fundamental machine learning concepts in the context of semantic search.
Given the rapid increase of interest in artificial intelligence in the last year, driven by many “generative transformer” technologies, it’s often hard to understand the foundational principles behind these advanced systems, especially for people not familiar with the mathematical theory.
We’ll use Ferdinand de Saussure’s theory of language as a starting point to gain a better understanding of these fundamental concepts, such as word embeddings, vectors, semantic similarity or multi-modal search, demystifying their “magic” without using complicated jargon or complex formulas.
Karel Minařík is a software developer with a focus on system architectures, web APIs, interactive applications, and data visualization.
For almost a decade, he has been working at Elastic, writing Ruby and Go code for libraries and internal applications. Until last year, he had been helping with the digital transformation in the Czech Republic as part of the Česko.Digital team.
