Factorized Databases

Explore the foundations and applications of factorized databases in this 51-minute lecture by Ahmet Kara from the University of Zurich. Delve into compressed yet lossless representations of relational data that enable efficient processing in the compressed domain. Learn about relational algebra expressions using union and Cartesian product operators, and how distributivity of product over union reduces redundancy in tabular data representation. Discover how factorized representations of join results can be computed directly from input databases, with time complexity proportional to their sizes. Gain insights into the decade-long progress in theory, systems, and applications of factorized databases. Examine their use in static and dynamic relational algorithms, probabilistic databases, and in-database machine learning.