My goal is to understand well a paper like ApproxMC. It discusses the use of Hash functions for Propositional Model Counting. In my understanding what they call hash functions are just random XOR's added to propositional formulas. On the other hand any tutorial level discussion on hashing is completely dedicated to putting data in bins for easy access which seems completely divorced from their discussion.
I would like to have a reference on randomised algorithms, which is able to explain universal hashing from scratch and is able to build up to the level of rigorous statistical understanding of universal hashing families.
Other similar problems in my understanding is LSH (Locality Sensitive Hashing).