Categorical Models for BigData

Abstract

This paper shows how concepts coming from category theory associated to a functional programming language can help to formalize and reason about data and get efficient programs in a BigData context. More precisely, it shows how data structures can be modeled by functors related by natural transformations (and isomorphisms). The transformation functions can then serve to shift a data structure and then get another program (eventually educing time complexity). The paper then explains the main concepts of the theory, how to apply them and gives an application to a concrete database and the performances obtained.