1: \begin{abstract}

2: Similarity caching allows requests for an item 

3:  to be served by a similar item. Applications include recommendation systems,

4: multimedia retrieval, and machine learning. Recently, many similarity

5: caching policies have been proposed, like SIM-LRU and

6: RND-LRU, but the performance analysis of their hit rate is still wanting.

7: In this paper, we show 

8: how to extend the popular time-to-live approximation in classic caching to similarity

9: caching. In particular, we propose a method to estimate the hit rate of the similarity caching policy RND-LRU. Our method, the RND-TTL approximation, introduces the RND-TTL cache model and then tunes its parameters in such a way to mimic the behavior of RND-LRU. The parameter tuning involves solving a fixed point system of equations for which we provide an algorithm for numerical resolution and sufficient conditions for its convergence. Our approach for approximating the hit rate of RND-LRU is evaluated on both synthetic and real world traces. 

10: \end{abstract}

11: