Random-Access Ranked Retrieval and Similarity Search

We extend Random Access, a fundamental operation that enables efficient search and exploration algorithms, to the modern interactive data systems based on Ranked Retrieval and Similarity Search, where orderings are dynamically defined over a high-dimensional feature space. This extension enables efficient solutions for a wide range of applications, from data analytics tools and database systems to recommendation systems and machine learning.